Daily Papers

In this paper, we target at the problem of learning a generalizable dynamic radiance field from monocular videos. Different from most existing NeRF methods that are based on multiple views, monocular videos only contain one view at each timestamp, thereby suffering from ambiguity along the view direction in estimating point features and scene flows. Previous studies such as DynNeRF disambiguate point features by positional encoding, which is not transferable and severely limits the generalization ability. As a result, these methods have to train one independent model for each scene and suffer from heavy computational costs when applying to increasing monocular videos in real-world applications. To address this, We propose MonoNeRF to simultaneously learn point features and scene flows with point trajectory and feature correspondence constraints across frames. More specifically, we learn an implicit velocity field to estimate point trajectory from temporal features with Neural ODE, which is followed by a flow-based feature aggregation module to obtain spatial features along the point trajectory. We jointly optimize temporal and spatial features in an end-to-end manner. Experiments show that our MonoNeRF is able to learn from multiple scenes and support new applications such as scene editing, unseen frame synthesis, and fast novel scene adaptation. Codes are available at https://github.com/tianfr/MonoNeRF.

We address the problem of synthesizing novel views from a monocular video depicting a complex dynamic scene. State-of-the-art methods based on temporally varying Neural Radiance Fields (aka dynamic NeRFs) have shown impressive results on this task. However, for long videos with complex object motions and uncontrolled camera trajectories, these methods can produce blurry or inaccurate renderings, hampering their use in real-world applications. Instead of encoding the entire dynamic scene within the weights of MLPs, we present a new approach that addresses these limitations by adopting a volumetric image-based rendering framework that synthesizes new viewpoints by aggregating features from nearby views in a scene-motion-aware manner. Our system retains the advantages of prior methods in its ability to model complex scenes and view-dependent effects, but also enables synthesizing photo-realistic novel views from long videos featuring complex scene dynamics with unconstrained camera trajectories. We demonstrate significant improvements over state-of-the-art methods on dynamic scene datasets, and also apply our approach to in-the-wild videos with challenging camera and object motion, where prior methods fail to produce high-quality renderings. Our project webpage is at dynibar.github.io.

Video generation models have demonstrated great capabilities of producing impressive monocular videos, however, the generation of 3D stereoscopic video remains under-explored. We propose a pose-free and training-free approach for generating 3D stereoscopic videos using an off-the-shelf monocular video generation model. Our method warps a generated monocular video into camera views on stereoscopic baseline using estimated video depth, and employs a novel frame matrix video inpainting framework. The framework leverages the video generation model to inpaint frames observed from different timestamps and views. This effective approach generates consistent and semantically coherent stereoscopic videos without scene optimization or model fine-tuning. Moreover, we develop a disocclusion boundary re-injection scheme that further improves the quality of video inpainting by alleviating the negative effects propagated from disoccluded areas in the latent space. We validate the efficacy of our proposed method by conducting experiments on videos from various generative models, including Sora [4 ], Lumiere [2], WALT [8 ], and Zeroscope [ 42]. The experiments demonstrate that our method has a significant improvement over previous methods. The code will be released at https://daipengwa.github.io/SVG_ProjectPage.

Self-supervised monocular depth estimation has gathered notable interest since it can liberate training from dependency on depth annotations. In monocular video training case, recent methods only conduct view synthesis between existing camera views, leading to insufficient guidance. To tackle this, we try to synthesize more virtual camera views by flow-based video frame interpolation (VFI), termed as temporal augmentation. For multi-frame inference, to sidestep the problem of dynamic objects encountered by explicit geometry-based methods like ManyDepth, we return to the feature fusion paradigm and design a VFI-assisted multi-frame fusion module to align and aggregate multi-frame features, using motion and occlusion information obtained by the flow-based VFI model. Finally, we construct a unified self-supervised learning framework, named Mono-ViFI, to bilaterally connect single- and multi-frame depth. In this framework, spatial data augmentation through image affine transformation is incorporated for data diversity, along with a triplet depth consistency loss for regularization. The single- and multi-frame models can share weights, making our framework compact and memory-efficient. Extensive experiments demonstrate that our method can bring significant improvements to current advanced architectures. Source code is available at https://github.com/LiuJF1226/Mono-ViFI.

In light of recent advances in multimodal Large Language Models (LLMs), there is increasing attention to scaling them from image-text data to more informative real-world videos. Compared to static images, video poses unique challenges for effective large-scale pre-training due to the modeling of its spatiotemporal dynamics. In this paper, we address such limitations in video-language pre-training with an efficient video decomposition that represents each video as keyframes and temporal motions. These are then adapted to an LLM using well-designed tokenizers that discretize visual and temporal information as a few tokens, thus enabling unified generative pre-training of videos, images, and text. At inference, the generated tokens from the LLM are carefully recovered to the original continuous pixel space to create various video content. Our proposed framework is both capable of comprehending and generating image and video content, as demonstrated by its competitive performance across 13 multimodal benchmarks in image and video understanding and generation. Our code and models will be available at https://video-lavit.github.io.

Depth Anything has achieved remarkable success in monocular depth estimation with strong generalization ability. However, it suffers from temporal inconsistency in videos, hindering its practical applications. Various methods have been proposed to alleviate this issue by leveraging video generation models or introducing priors from optical flow and camera poses. Nonetheless, these methods are only applicable to short videos (< 10 seconds) and require a trade-off between quality and computational efficiency. We propose Video Depth Anything for high-quality, consistent depth estimation in super-long videos (over several minutes) without sacrificing efficiency. We base our model on Depth Anything V2 and replace its head with an efficient spatial-temporal head. We design a straightforward yet effective temporal consistency loss by constraining the temporal depth gradient, eliminating the need for additional geometric priors. The model is trained on a joint dataset of video depth and unlabeled images, similar to Depth Anything V2. Moreover, a novel key-frame-based strategy is developed for long video inference. Experiments show that our model can be applied to arbitrarily long videos without compromising quality, consistency, or generalization ability. Comprehensive evaluations on multiple video benchmarks demonstrate that our approach sets a new state-of-the-art in zero-shot video depth estimation. We offer models of different scales to support a range of scenarios, with our smallest model capable of real-time performance at 30 FPS.

We present a system that allows for accurate, fast, and robust estimation of camera parameters and depth maps from casual monocular videos of dynamic scenes. Most conventional structure from motion and monocular SLAM techniques assume input videos that feature predominantly static scenes with large amounts of parallax. Such methods tend to produce erroneous estimates in the absence of these conditions. Recent neural network-based approaches attempt to overcome these challenges; however, such methods are either computationally expensive or brittle when run on dynamic videos with uncontrolled camera motion or unknown field of view. We demonstrate the surprising effectiveness of a deep visual SLAM framework: with careful modifications to its training and inference schemes, this system can scale to real-world videos of complex dynamic scenes with unconstrained camera paths, including videos with little camera parallax. Extensive experiments on both synthetic and real videos demonstrate that our system is significantly more accurate and robust at camera pose and depth estimation when compared with prior and concurrent work, with faster or comparable running times. See interactive results on our project page: https://mega-sam.github.io/

Large-scale pretrained image-text models have shown incredible zero-shot performance in a handful of tasks, including video ones such as action recognition and text-to-video retrieval. However, these models have not been adapted to video, mainly because they do not account for the time dimension but also because video frames are different from the typical images (e.g., containing motion blur, and less sharpness). In this paper, we present a fine-tuning strategy to refine these large-scale pretrained image-text models for zero-shot video understanding tasks. We show that by carefully adapting these models we obtain considerable improvements on two zero-shot Action Recognition tasks and three zero-shot Text-to-video Retrieval tasks. The code is available at https://github.com/bryant1410/fitclip

Accurate reconstruction of complex dynamic scenes from just a single viewpoint continues to be a challenging task in computer vision. Current dynamic novel view synthesis methods typically require videos from many different camera viewpoints, necessitating careful recording setups, and significantly restricting their utility in the wild as well as in terms of embodied AI applications. In this paper, we propose GCD, a controllable monocular dynamic view synthesis pipeline that leverages large-scale diffusion priors to, given a video of any scene, generate a synchronous video from any other chosen perspective, conditioned on a set of relative camera pose parameters. Our model does not require depth as input, and does not explicitly model 3D scene geometry, instead performing end-to-end video-to-video translation in order to achieve its goal efficiently. Despite being trained on synthetic multi-view video data only, zero-shot real-world generalization experiments show promising results in multiple domains, including robotics, object permanence, and driving environments. We believe our framework can potentially unlock powerful applications in rich dynamic scene understanding, perception for robotics, and interactive 3D video viewing experiences for virtual reality.

Monocular dynamic reconstruction is a challenging and long-standing vision problem due to the highly ill-posed nature of the task. Existing approaches are limited in that they either depend on templates, are effective only in quasi-static scenes, or fail to model 3D motion explicitly. In this work, we introduce a method capable of reconstructing generic dynamic scenes, featuring explicit, full-sequence-long 3D motion, from casually captured monocular videos. We tackle the under-constrained nature of the problem with two key insights: First, we exploit the low-dimensional structure of 3D motion by representing scene motion with a compact set of SE3 motion bases. Each point's motion is expressed as a linear combination of these bases, facilitating soft decomposition of the scene into multiple rigidly-moving groups. Second, we utilize a comprehensive set of data-driven priors, including monocular depth maps and long-range 2D tracks, and devise a method to effectively consolidate these noisy supervisory signals, resulting in a globally consistent representation of the dynamic scene. Experiments show that our method achieves state-of-the-art performance for both long-range 3D/2D motion estimation and novel view synthesis on dynamic scenes. Project Page: https://shape-of-motion.github.io/

Novel view synthesis from an in-the-wild video is difficult due to challenges like scene dynamics and lack of parallax. While existing methods have shown promising results with implicit neural radiance fields, they are slow to train and render. This paper revisits explicit video representations to synthesize high-quality novel views from a monocular video efficiently. We treat static and dynamic video content separately. Specifically, we build a global static scene model using an extended plane-based scene representation to synthesize temporally coherent novel video. Our plane-based scene representation is augmented with spherical harmonics and displacement maps to capture view-dependent effects and model non-planar complex surface geometry. We opt to represent the dynamic content as per-frame point clouds for efficiency. While such representations are inconsistency-prone, minor temporal inconsistencies are perceptually masked due to motion. We develop a method to quickly estimate such a hybrid video representation and render novel views in real time. Our experiments show that our method can render high-quality novel views from an in-the-wild video with comparable quality to state-of-the-art methods while being 100x faster in training and enabling real-time rendering.

Self-supervised learning has unlocked the potential of scaling up pretraining to billions of images, since annotation is unnecessary. But are we making the best use of data? How more economical can we be? In this work, we attempt to answer this question by making two contributions. First, we investigate first-person videos and introduce a "Walking Tours" dataset. These videos are high-resolution, hours-long, captured in a single uninterrupted take, depicting a large number of objects and actions with natural scene transitions. They are unlabeled and uncurated, thus realistic for self-supervision and comparable with human learning. Second, we introduce a novel self-supervised image pretraining method tailored for learning from continuous videos. Existing methods typically adapt image-based pretraining approaches to incorporate more frames. Instead, we advocate a "tracking to learn to recognize" approach. Our method called DoRA, leads to attention maps that Discover and tRAck objects over time in an end-to-end manner, using transformer cross-attention. We derive multiple views from the tracks and use them in a classical self-supervised distillation loss. Using our novel approach, a single Walking Tours video remarkably becomes a strong competitor to ImageNet for several image and video downstream tasks.

We propose a generalizable neural radiance fields - MonoNeRF, that can be trained on large-scale monocular videos of moving in static scenes without any ground-truth annotations of depth and camera poses. MonoNeRF follows an Autoencoder-based architecture, where the encoder estimates the monocular depth and the camera pose, and the decoder constructs a Multiplane NeRF representation based on the depth encoder feature, and renders the input frames with the estimated camera. The learning is supervised by the reconstruction error. Once the model is learned, it can be applied to multiple applications including depth estimation, camera pose estimation, and single-image novel view synthesis. More qualitative results are available at: https://oasisyang.github.io/mononerf .

We present TrajectoryCrafter, a novel approach to redirect camera trajectories for monocular videos. By disentangling deterministic view transformations from stochastic content generation, our method achieves precise control over user-specified camera trajectories. We propose a novel dual-stream conditional video diffusion model that concurrently integrates point cloud renders and source videos as conditions, ensuring accurate view transformations and coherent 4D content generation. Instead of leveraging scarce multi-view videos, we curate a hybrid training dataset combining web-scale monocular videos with static multi-view datasets, by our innovative double-reprojection strategy, significantly fostering robust generalization across diverse scenes. Extensive evaluations on multi-view and large-scale monocular videos demonstrate the superior performance of our method.

The increasing variety and quantity of tagged multimedia content on a variety of online platforms offer a unique opportunity to advance the field of human action recognition. In this study, we utilize 283,582 unique, unlabeled TikTok video clips, categorized into 386 hashtags, to train a domain-specific foundation model for action recognition. We employ VideoMAE V2, an advanced model integrating Masked Autoencoders (MAE) with Vision Transformers (ViT), pre-trained on this diverse collection of unstructured videos. Our model, fine-tuned on established action recognition benchmarks such as UCF101 and HMDB51, achieves state-of-the-art results: 99.05% on UCF101, 86.08% on HMDB51, 85.51% on Kinetics-400, and 74.27% on Something-Something V2 using the ViT-giant backbone. These results highlight the potential of using unstructured and unlabeled videos as a valuable source of diverse and dynamic content for training foundation models. Our investigation confirms that while initial increases in pre-training data volume significantly enhance model performance, the gains diminish as the dataset size continues to expand. Our findings emphasize two critical axioms in self-supervised learning for computer vision: (1) additional pre-training data can yield diminishing benefits for some datasets and (2) quality is more important than quantity in self-supervised learning, especially when building foundation models.

Synthesizing novel views from in-the-wild monocular videos is challenging due to scene dynamics and the lack of multi-view cues. To address this, we propose SplineGS, a COLMAP-free dynamic 3D Gaussian Splatting (3DGS) framework for high-quality reconstruction and fast rendering from monocular videos. At its core is a novel Motion-Adaptive Spline (MAS) method, which represents continuous dynamic 3D Gaussian trajectories using cubic Hermite splines with a small number of control points. For MAS, we introduce a Motion-Adaptive Control points Pruning (MACP) method to model the deformation of each dynamic 3D Gaussian across varying motions, progressively pruning control points while maintaining dynamic modeling integrity. Additionally, we present a joint optimization strategy for camera parameter estimation and 3D Gaussian attributes, leveraging photometric and geometric consistency. This eliminates the need for Structure-from-Motion preprocessing and enhances SplineGS's robustness in real-world conditions. Experiments show that SplineGS significantly outperforms state-of-the-art methods in novel view synthesis quality for dynamic scenes from monocular videos, achieving thousands times faster rendering speed.

In this paper, we show that useful video representations can be learned from synthetic videos and natural images, without incorporating natural videos in the training. We propose a progression of video datasets synthesized by simple generative processes, that model a growing set of natural video properties (e.g. motion, acceleration, and shape transformations). The downstream performance of video models pre-trained on these generated datasets gradually increases with the dataset progression. A VideoMAE model pre-trained on our synthetic videos closes 97.2% of the performance gap on UCF101 action classification between training from scratch and self-supervised pre-training from natural videos, and outperforms the pre-trained model on HMDB51. Introducing crops of static images to the pre-training stage results in similar performance to UCF101 pre-training and outperforms the UCF101 pre-trained model on 11 out of 14 out-of-distribution datasets of UCF101-P. Analyzing the low-level properties of the datasets, we identify correlations between frame diversity, frame similarity to natural data, and downstream performance. Our approach provides a more controllable and transparent alternative to video data curation processes for pre-training.

Recent years have seen a tremendous improvement in the quality of video generation and editing approaches. While several techniques focus on editing appearance, few address motion. Current approaches using text, trajectories, or bounding boxes are limited to simple motions, so we specify motions with a single motion reference video instead. We further propose to use a pre-trained image-to-video model rather than a text-to-video model. This approach allows us to preserve the exact appearance and position of a target object or scene and helps disentangle appearance from motion. Our method, called motion-textual inversion, leverages our observation that image-to-video models extract appearance mainly from the (latent) image input, while the text/image embedding injected via cross-attention predominantly controls motion. We thus represent motion using text/image embedding tokens. By operating on an inflated motion-text embedding containing multiple text/image embedding tokens per frame, we achieve a high temporal motion granularity. Once optimized on the motion reference video, this embedding can be applied to various target images to generate videos with semantically similar motions. Our approach does not require spatial alignment between the motion reference video and target image, generalizes across various domains, and can be applied to various tasks such as full-body and face reenactment, as well as controlling the motion of inanimate objects and the camera. We empirically demonstrate the effectiveness of our method in the semantic video motion transfer task, significantly outperforming existing methods in this context.

This paper presents a unified multimodal pre-trained model called N\"UWA that can generate new or manipulate existing visual data (i.e., images and videos) for various visual synthesis tasks. To cover language, image, and video at the same time for different scenarios, a 3D transformer encoder-decoder framework is designed, which can not only deal with videos as 3D data but also adapt to texts and images as 1D and 2D data, respectively. A 3D Nearby Attention (3DNA) mechanism is also proposed to consider the nature of the visual data and reduce the computational complexity. We evaluate N\"UWA on 8 downstream tasks. Compared to several strong baselines, N\"UWA achieves state-of-the-art results on text-to-image generation, text-to-video generation, video prediction, etc. Furthermore, it also shows surprisingly good zero-shot capabilities on text-guided image and video manipulation tasks. Project repo is https://github.com/microsoft/NUWA.

With the explosive growth of web videos and emerging large-scale vision-language pre-training models, e.g., CLIP, retrieving videos of interest with text instructions has attracted increasing attention. A common practice is to transfer text-video pairs to the same embedding space and craft cross-modal interactions with certain entities in specific granularities for semantic correspondence. Unfortunately, the intrinsic uncertainties of optimal entity combinations in appropriate granularities for cross-modal queries are understudied, which is especially critical for modalities with hierarchical semantics, e.g., video, text, etc. In this paper, we propose an Uncertainty-Adaptive Text-Video Retrieval approach, termed UATVR, which models each look-up as a distribution matching procedure. Concretely, we add additional learnable tokens in the encoders to adaptively aggregate multi-grained semantics for flexible high-level reasoning. In the refined embedding space, we represent text-video pairs as probabilistic distributions where prototypes are sampled for matching evaluation. Comprehensive experiments on four benchmarks justify the superiority of our UATVR, which achieves new state-of-the-art results on MSR-VTT (50.8%), VATEX (64.5%), MSVD (49.7%), and DiDeMo (45.8%). The code is available at https://github.com/bofang98/UATVR.

We present a real-time monocular dense SLAM system designed bottom-up from MASt3R, a two-view 3D reconstruction and matching prior. Equipped with this strong prior, our system is robust on in-the-wild video sequences despite making no assumption on a fixed or parametric camera model beyond a unique camera centre. We introduce efficient methods for pointmap matching, camera tracking and local fusion, graph construction and loop closure, and second-order global optimisation. With known calibration, a simple modification to the system achieves state-of-the-art performance across various benchmarks. Altogether, we propose a plug-and-play monocular SLAM system capable of producing globally-consistent poses and dense geometry while operating at 15 FPS.

We address the problem of generating videos from unposed internet photos. A handful of input images serve as keyframes, and our model interpolates between them to simulate a path moving between the cameras. Given random images, a model's ability to capture underlying geometry, recognize scene identity, and relate frames in terms of camera position and orientation reflects a fundamental understanding of 3D structure and scene layout. However, existing video models such as Luma Dream Machine fail at this task. We design a self-supervised method that takes advantage of the consistency of videos and variability of multiview internet photos to train a scalable, 3D-aware video model without any 3D annotations such as camera parameters. We validate that our method outperforms all baselines in terms of geometric and appearance consistency. We also show our model benefits applications that enable camera control, such as 3D Gaussian Splatting. Our results suggest that we can scale up scene-level 3D learning using only 2D data such as videos and multiview internet photos.

This paper introduces MiniGPT4-Video, a multimodal Large Language Model (LLM) designed specifically for video understanding. The model is capable of processing both temporal visual and textual data, making it adept at understanding the complexities of videos. Building upon the success of MiniGPT-v2, which excelled in translating visual features into the LLM space for single images and achieved impressive results on various image-text benchmarks, this paper extends the model's capabilities to process a sequence of frames, enabling it to comprehend videos. MiniGPT4-video does not only consider visual content but also incorporates textual conversations, allowing the model to effectively answer queries involving both visual and text components. The proposed model outperforms existing state-of-the-art methods, registering gains of 4.22%, 1.13%, 20.82%, and 13.1% on the MSVD, MSRVTT, TGIF, and TVQA benchmarks respectively. Our models and code have been made publicly available here https://vision-cair.github.io/MiniGPT4-video/

Stereo video synthesis from a monocular input is a demanding task in the fields of spatial computing and virtual reality. The main challenges of this task lie on the insufficiency of high-quality paired stereo videos for training and the difficulty of maintaining the spatio-temporal consistency between frames. Existing methods primarily address these issues by directly applying novel view synthesis (NVS) techniques to video, while facing limitations such as the inability to effectively represent dynamic scenes and the requirement for large amounts of training data. In this paper, we introduce a novel self-supervised stereo video synthesis paradigm via a video diffusion model, termed SpatialDreamer, which meets the challenges head-on. Firstly, to address the stereo video data insufficiency, we propose a Depth based Video Generation module DVG, which employs a forward-backward rendering mechanism to generate paired videos with geometric and temporal priors. Leveraging data generated by DVG, we propose RefinerNet along with a self-supervised synthetic framework designed to facilitate efficient and dedicated training. More importantly, we devise a consistency control module, which consists of a metric of stereo deviation strength and a Temporal Interaction Learning module TIL for geometric and temporal consistency ensurance respectively. We evaluated the proposed method against various benchmark methods, with the results showcasing its superior performance.

We present an algorithm for reconstructing dense, geometrically consistent depth for all pixels in a monocular video. We leverage a conventional structure-from-motion reconstruction to establish geometric constraints on pixels in the video. Unlike the ad-hoc priors in classical reconstruction, we use a learning-based prior, i.e., a convolutional neural network trained for single-image depth estimation. At test time, we fine-tune this network to satisfy the geometric constraints of a particular input video, while retaining its ability to synthesize plausible depth details in parts of the video that are less constrained. We show through quantitative validation that our method achieves higher accuracy and a higher degree of geometric consistency than previous monocular reconstruction methods. Visually, our results appear more stable. Our algorithm is able to handle challenging hand-held captured input videos with a moderate degree of dynamic motion. The improved quality of the reconstruction enables several applications, such as scene reconstruction and advanced video-based visual effects.

We present the TinyLLaVA-Video, a video understanding model with parameters not exceeding 4B that processes video sequences in a simple manner, without the need for complex architectures, supporting both fps sampling and uniform frame sampling. Our model is characterized by modularity and scalability, allowing training and inference with limited computational resources and enabling users to replace components based on their needs. We validate the effectiveness of this framework through experiments, the best model achieving performance comparable to certain existing 7B models on multiple video understanding benchmarks. The code and training recipes are fully open source, with all components and training data publicly available. We hope this work can serve as a baseline for practitioners exploring small-scale multimodal models for video understanding. It is available at https://github.com/ZhangXJ199/TinyLLaVA-Video.

Rendering scenes observed in a monocular video from novel viewpoints is a challenging problem. For static scenes the community has studied both scene-specific optimization techniques, which optimize on every test scene, and generalized techniques, which only run a deep net forward pass on a test scene. In contrast, for dynamic scenes, scene-specific optimization techniques exist, but, to our best knowledge, there is currently no generalized method for dynamic novel view synthesis from a given monocular video. To answer whether generalized dynamic novel view synthesis from monocular videos is possible today, we establish an analysis framework based on existing techniques and work toward the generalized approach. We find a pseudo-generalized process without scene-specific appearance optimization is possible, but geometrically and temporally consistent depth estimates are needed. Despite no scene-specific appearance optimization, the pseudo-generalized approach improves upon some scene-specific methods.

We extend neural radiance fields (NeRFs) to dynamic large-scale urban scenes. Prior work tends to reconstruct single video clips of short durations (up to 10 seconds). Two reasons are that such methods (a) tend to scale linearly with the number of moving objects and input videos because a separate model is built for each and (b) tend to require supervision via 3D bounding boxes and panoptic labels, obtained manually or via category-specific models. As a step towards truly open-world reconstructions of dynamic cities, we introduce two key innovations: (a) we factorize the scene into three separate hash table data structures to efficiently encode static, dynamic, and far-field radiance fields, and (b) we make use of unlabeled target signals consisting of RGB images, sparse LiDAR, off-the-shelf self-supervised 2D descriptors, and most importantly, 2D optical flow. Operationalizing such inputs via photometric, geometric, and feature-metric reconstruction losses enables SUDS to decompose dynamic scenes into the static background, individual objects, and their motions. When combined with our multi-branch table representation, such reconstructions can be scaled to tens of thousands of objects across 1.2 million frames from 1700 videos spanning geospatial footprints of hundreds of kilometers, (to our knowledge) the largest dynamic NeRF built to date. We present qualitative initial results on a variety of tasks enabled by our representations, including novel-view synthesis of dynamic urban scenes, unsupervised 3D instance segmentation, and unsupervised 3D cuboid detection. To compare to prior work, we also evaluate on KITTI and Virtual KITTI 2, surpassing state-of-the-art methods that rely on ground truth 3D bounding box annotations while being 10x quicker to train.

Image-based visual-language (I-VL) pre-training has shown great success for learning joint visual-textual representations from large-scale web data, revealing remarkable ability for zero-shot generalisation. This paper presents a simple but strong baseline to efficiently adapt the pre-trained I-VL model, and exploit its powerful ability for resource-hungry video understanding tasks, with minimal training. Specifically, we propose to optimise a few random vectors, termed as continuous prompt vectors, that convert video-related tasks into the same format as the pre-training objectives. In addition, to bridge the gap between static images and videos, temporal information is encoded with lightweight Transformers stacking on top of frame-wise visual features. Experimentally, we conduct extensive ablation studies to analyse the critical components. On 10 public benchmarks of action recognition, action localisation, and text-video retrieval, across closed-set, few-shot, and zero-shot scenarios, we achieve competitive or state-of-the-art performance to existing methods, despite optimising significantly fewer parameters.

The correlation between the vision and text is essential for video moment retrieval (VMR), however, existing methods heavily rely on separate pre-training feature extractors for visual and textual understanding. Without sufficient temporal boundary annotations, it is non-trivial to learn universal video-text alignments. In this work, we explore multi-modal correlations derived from large-scale image-text data to facilitate generalisable VMR. To address the limitations of image-text pre-training models on capturing the video changes, we propose a generic method, referred to as Visual-Dynamic Injection (VDI), to empower the model's understanding of video moments. Whilst existing VMR methods are focusing on building temporal-aware video features, being aware of the text descriptions about the temporal changes is also critical but originally overlooked in pre-training by matching static images with sentences. Therefore, we extract visual context and spatial dynamic information from video frames and explicitly enforce their alignments with the phrases describing video changes (e.g. verb). By doing so, the potentially relevant visual and motion patterns in videos are encoded in the corresponding text embeddings (injected) so to enable more accurate video-text alignments. We conduct extensive experiments on two VMR benchmark datasets (Charades-STA and ActivityNet-Captions) and achieve state-of-the-art performances. Especially, VDI yields notable advantages when being tested on the out-of-distribution splits where the testing samples involve novel scenes and vocabulary.

Learning specific hands-on skills such as cooking, car maintenance, and home repairs increasingly happens via instructional videos. The user experience with such videos is known to be improved by meta-information such as time-stamped annotations for the main steps involved. Generating such annotations automatically is challenging, and we describe here two relevant contributions. First, we construct and release a new dense video captioning dataset, Video Timeline Tags (ViTT), featuring a variety of instructional videos together with time-stamped annotations. Second, we explore several multimodal sequence-to-sequence pretraining strategies that leverage large unsupervised datasets of videos and caption-like texts. We pretrain and subsequently finetune dense video captioning models using both YouCook2 and ViTT. We show that such models generalize well and are robust over a wide variety of instructional videos.

Image generative models have made significant progress in generating realistic and diverse images, supported by comprehensive guidance from various evaluation metrics. However, current video generative models struggle to generate even short video clips, with limited tools that provide insights for improvements. Current video evaluation metrics are simple adaptations of image metrics by switching the embeddings with video embedding networks, which may underestimate the unique characteristics of video. Our analysis reveals that the widely used Frechet Video Distance (FVD) has a stronger emphasis on the spatial aspect than the temporal naturalness of video and is inherently constrained by the input size of the embedding networks used, limiting it to 16 frames. Additionally, it demonstrates considerable instability and diverges from human evaluations. To address the limitations, we propose STREAM, a new video evaluation metric uniquely designed to independently evaluate spatial and temporal aspects. This feature allows comprehensive analysis and evaluation of video generative models from various perspectives, unconstrained by video length. We provide analytical and experimental evidence demonstrating that STREAM provides an effective evaluation tool for both visual and temporal quality of videos, offering insights into area of improvement for video generative models. To the best of our knowledge, STREAM is the first evaluation metric that can separately assess the temporal and spatial aspects of videos. Our code is available at https://github.com/pro2nit/STREAM.

Mainstream Video-Language Pre-training models actbert,clipbert,violet consist of three parts, a video encoder, a text encoder, and a video-text fusion Transformer. They pursue better performance via utilizing heavier unimodal encoders or multimodal fusion Transformers, resulting in increased parameters with lower efficiency in downstream tasks. In this work, we for the first time introduce an end-to-end video-language model, namely all-in-one Transformer, that embeds raw video and textual signals into joint representations using a unified backbone architecture. We argue that the unique temporal information of video data turns out to be a key barrier hindering the design of a modality-agnostic Transformer. To overcome the challenge, we introduce a novel and effective token rolling operation to encode temporal representations from video clips in a non-parametric manner. The careful design enables the representation learning of both video-text multimodal inputs and unimodal inputs using a unified backbone model. Our pre-trained all-in-one Transformer is transferred to various downstream video-text tasks after fine-tuning, including text-video retrieval, video-question answering, multiple choice and visual commonsense reasoning. State-of-the-art performances with the minimal model FLOPs on nine datasets demonstrate the superiority of our method compared to the competitive counterparts. The code and pretrained model have been released in https://github.com/showlab/all-in-one.

Existing deep video models are limited by specific tasks, fixed input-output spaces, and poor generalization capabilities, making it difficult to deploy them in real-world scenarios. In this paper, we present our vision for multimodal and versatile video understanding and propose a prototype system, \system. Our system is built upon a tracklet-centric paradigm, which treats tracklets as the basic video unit and employs various Video Foundation Models (ViFMs) to annotate their properties e.g., appearance, motion, \etc. All the detected tracklets are stored in a database and interact with the user through a database manager. We have conducted extensive case studies on different types of in-the-wild videos, which demonstrates the effectiveness of our method in answering various video-related problems. Our project is available at https://www.wangjunke.info/ChatVideo/

A great challenge in video-language (VidL) modeling lies in the disconnection between fixed video representations extracted from image/video understanding models and downstream VidL data. Recent studies try to mitigate this disconnection via end-to-end training. To make it computationally feasible, prior works tend to "imagify" video inputs, i.e., a handful of sparsely sampled frames are fed into a 2D CNN, followed by a simple mean-pooling or concatenation to obtain the overall video representations. Although achieving promising results, such simple approaches may lose temporal information that is essential for performing downstream VidL tasks. In this work, we present VIOLET, a fully end-to-end VIdeO-LanguagE Transformer, which adopts a video transformer to explicitly model the temporal dynamics of video inputs. Further, unlike previous studies that found pre-training tasks on video inputs (e.g., masked frame modeling) not very effective, we design a new pre-training task, Masked Visual-token Modeling (MVM), for better video modeling. Specifically, the original video frame patches are "tokenized" into discrete visual tokens, and the goal is to recover the original visual tokens based on the masked patches. Comprehensive analysis demonstrates the effectiveness of both explicit temporal modeling via video transformer and MVM. As a result, VIOLET achieves new state-of-the-art performance on 5 video question answering tasks and 4 text-to-video retrieval tasks.

Large Vision-Language Models (VLMs) have been extended to understand both images and videos. Visual token compression is leveraged to reduce the considerable token length of visual inputs. To meet the needs of different tasks, existing high-performance models usually process images and videos separately with different token compression strategies, limiting the capabilities of combining images and videos. To this end, we extend each image into a "static" video and introduce a unified token compression strategy called Progressive Visual Token Compression (PVC), where the tokens of each frame are progressively encoded and adaptively compressed to supplement the information not extracted from previous frames. Video tokens are efficiently compressed with exploiting the inherent temporal redundancy. Images are repeated as static videos, and the spatial details can be gradually supplemented in multiple frames. PVC unifies the token compressing of images and videos. With a limited number of tokens per frame (64 tokens by default), spatial details and temporal changes can still be preserved. Experiments show that our model achieves state-of-the-art performance across various video understanding benchmarks, including long video tasks and fine-grained short video tasks. Meanwhile, our unified token compression strategy incurs no performance loss on image benchmarks, particularly in detail-sensitive tasks.

We release MiDaS v3.1 for monocular depth estimation, offering a variety of new models based on different encoder backbones. This release is motivated by the success of transformers in computer vision, with a large variety of pretrained vision transformers now available. We explore how using the most promising vision transformers as image encoders impacts depth estimation quality and runtime of the MiDaS architecture. Our investigation also includes recent convolutional approaches that achieve comparable quality to vision transformers in image classification tasks. While the previous release MiDaS v3.0 solely leverages the vanilla vision transformer ViT, MiDaS v3.1 offers additional models based on BEiT, Swin, SwinV2, Next-ViT and LeViT. These models offer different performance-runtime tradeoffs. The best model improves the depth estimation quality by 28% while efficient models enable downstream tasks requiring high frame rates. We also describe the general process for integrating new backbones. A video summarizing the work can be found at https://youtu.be/UjaeNNFf9sE and the code is available at https://github.com/isl-org/MiDaS.

Recent developments in monocular depth estimation methods enable high-quality depth estimation of single-view images but fail to estimate consistent video depth across different frames. Recent works address this problem by applying a video diffusion model to generate video depth conditioned on the input video, which is training-expensive and can only produce scale-invariant depth values without camera poses. In this paper, we propose a novel video-depth estimation method called Align3R to estimate temporal consistent depth maps for a dynamic video. Our key idea is to utilize the recent DUSt3R model to align estimated monocular depth maps of different timesteps. First, we fine-tune the DUSt3R model with additional estimated monocular depth as inputs for the dynamic scenes. Then, we apply optimization to reconstruct both depth maps and camera poses. Extensive experiments demonstrate that Align3R estimates consistent video depth and camera poses for a monocular video with superior performance than baseline methods.

Video-language pre-training has advanced the performance of various downstream video-language tasks. However, most previous methods directly inherit or adapt typical image-language pre-training paradigms to video-language pre-training, thus not fully exploiting the unique characteristic of video, i.e., temporal. In this paper, we propose a Hierarchical Temporal-Aware video-language pre-training framework, HiTeA, with two novel pre-training tasks for modeling cross-modal alignment between moments and texts as well as the temporal relations of video-text pairs. Specifically, we propose a cross-modal moment exploration task to explore moments in videos, which results in detailed video moment representation. Besides, the inherent temporal relations are captured by aligning video-text pairs as a whole in different time resolutions with multi-modal temporal relation exploration task. Furthermore, we introduce the shuffling test to evaluate the temporal reliance of datasets and video-language pre-training models. We achieve state-of-the-art results on 15 well-established video-language understanding and generation tasks, especially on temporal-oriented datasets (e.g., SSv2-Template and SSv2-Label) with 8.6% and 11.1% improvement respectively. HiTeA also demonstrates strong generalization ability when directly transferred to downstream tasks in a zero-shot manner. Models and demo will be available on ModelScope.

We present VideoFactory, an innovative framework for generating high-quality open-domain videos. VideoFactory excels in producing high-definition (1376x768), widescreen (16:9) videos without watermarks, creating an engaging user experience. Generating videos guided by text instructions poses significant challenges, such as modeling the complex relationship between space and time, and the lack of large-scale text-video paired data. Previous approaches extend pretrained text-to-image generation models by adding temporal 1D convolution/attention modules for video generation. However, these approaches overlook the importance of jointly modeling space and time, inevitably leading to temporal distortions and misalignment between texts and videos. In this paper, we propose a novel approach that strengthens the interaction between spatial and temporal perceptions. In particular, we utilize a swapped cross-attention mechanism in 3D windows that alternates the "query" role between spatial and temporal blocks, enabling mutual reinforcement for each other. To fully unlock model capabilities for high-quality video generation, we curate a large-scale video dataset called HD-VG-130M. This dataset comprises 130 million text-video pairs from the open-domain, ensuring high-definition, widescreen and watermark-free characters. Objective metrics and user studies demonstrate the superiority of our approach in terms of per-frame quality, temporal correlation, and text-video alignment, with clear margins.

With the widespread use of mobile devices and the rapid growth of micro-video platforms such as TikTok and Kwai, the demand for personalized micro-video recommendation systems has significantly increased. Micro-videos typically contain diverse information, such as textual metadata, visual cues (e.g., cover images), and dynamic video content, significantly affecting user interaction and engagement patterns. However, most existing approaches often suffer from the problem of over-smoothing, which limits their ability to capture comprehensive interaction information effectively. Additionally, cold-start scenarios present ongoing challenges due to sparse interaction data and the underutilization of available interaction signals. To address these issues, we propose a Multi-view Hypergraph-based Contrastive learning model for cold-start micro-video Recommendation (MHCR). MHCR introduces a multi-view multimodal feature extraction layer to capture interaction signals from various perspectives and incorporates multi-view self-supervised learning tasks to provide additional supervisory signals. Through extensive experiments on two real-world datasets, we show that MHCR significantly outperforms existing video recommendation models and effectively mitigates cold-start challenges. Our code is available at https://github.com/sisuolv/MHCR.

Recent advances in internet-scale video data pretraining have led to the development of text-to-video generative models that can create high-quality videos across a broad range of visual concepts, synthesize realistic motions and render complex objects. Hence, these generative models have the potential to become general-purpose simulators of the physical world. However, it is unclear how far we are from this goal with the existing text-to-video generative models. To this end, we present VideoPhy, a benchmark designed to assess whether the generated videos follow physical commonsense for real-world activities (e.g. marbles will roll down when placed on a slanted surface). Specifically, we curate diverse prompts that involve interactions between various material types in the physical world (e.g., solid-solid, solid-fluid, fluid-fluid). We then generate videos conditioned on these captions from diverse state-of-the-art text-to-video generative models, including open models (e.g., CogVideoX) and closed models (e.g., Lumiere, Dream Machine). Our human evaluation reveals that the existing models severely lack the ability to generate videos adhering to the given text prompts, while also lack physical commonsense. Specifically, the best performing model, CogVideoX-5B, generates videos that adhere to the caption and physical laws for 39.6% of the instances. VideoPhy thus highlights that the video generative models are far from accurately simulating the physical world. Finally, we propose an auto-evaluator, VideoCon-Physics, to assess the performance reliably for the newly released models.

This paper proposes a novel pretext task to address the self-supervised video representation learning problem. Specifically, given an unlabeled video clip, we compute a series of spatio-temporal statistical summaries, such as the spatial location and dominant direction of the largest motion, the spatial location and dominant color of the largest color diversity along the temporal axis, etc. Then a neural network is built and trained to yield the statistical summaries given the video frames as inputs. In order to alleviate the learning difficulty, we employ several spatial partitioning patterns to encode rough spatial locations instead of exact spatial Cartesian coordinates. Our approach is inspired by the observation that human visual system is sensitive to rapidly changing contents in the visual field, and only needs impressions about rough spatial locations to understand the visual contents. To validate the effectiveness of the proposed approach, we conduct extensive experiments with four 3D backbone networks, i.e., C3D, 3D-ResNet, R(2+1)D and S3D-G. The results show that our approach outperforms the existing approaches across these backbone networks on four downstream video analysis tasks including action recognition, video retrieval, dynamic scene recognition, and action similarity labeling. The source code is publicly available at: https://github.com/laura-wang/video_repres_sts.

We present VideoOrion, a Video Large Language Model (Video-LLM) that explicitly captures the key semantic information in videos--the spatial-temporal dynamics of objects throughout the videos. VideoOrion employs expert vision models to extract object dynamics through a detect-segment-track pipeline, encoding them into a set of object tokens by aggregating spatial-temporal object features. Our method addresses the persistent challenge in Video-LLMs of efficiently compressing high-dimensional video data into semantic tokens that are comprehensible to LLMs. Compared to prior methods which resort to downsampling the original video or aggregating visual tokens using resamplers, leading to information loss and entangled semantics, VideoOrion not only offers a more natural and efficient way to derive compact, disentangled semantic representations but also enables explicit object modeling of video content with minimal computational cost. Moreover, the introduced object tokens naturally allow VideoOrion to accomplish video-based referring tasks. Experimental results show that VideoOrion can learn to make good use of the object tokens, and achieves competitive results on both general video question answering and video-based referring benchmarks.

Large-scale video-language pre-training has made remarkable strides in advancing video-language understanding tasks. However, the heavy computational burden of video encoding remains a formidable efficiency bottleneck, particularly for long-form videos. These videos contain massive visual tokens due to their inherent 3D properties and spatiotemporal redundancy, making it challenging to capture complex temporal and spatial relationships. To tackle this issue, we propose an efficient method called TEmporal-Spatial Token Aggregation (TESTA). TESTA condenses video semantics by adaptively aggregating similar frames, as well as similar patches within each frame. TESTA can reduce the number of visual tokens by 75% and thus accelerate video encoding. Building upon TESTA, we introduce a pre-trained video-language model equipped with a divided space-time token aggregation module in each video encoder block. We evaluate our model on five datasets for paragraph-to-video retrieval and long-form VideoQA tasks. Experimental results show that TESTA improves computing efficiency by 1.7 times, and achieves significant performance gains from its scalability in processing longer input frames, e.g., +13.7 R@1 on QuerYD and +6.5 R@1 on Condensed Movie.

This paper presents a novel approach to Zero-Shot Action Recognition. Recent works have explored the detection and classification of objects to obtain semantic information from videos with remarkable performance. Inspired by them, we propose using video captioning methods to extract semantic information about objects, scenes, humans, and their relationships. To the best of our knowledge, this is the first work to represent both videos and labels with descriptive sentences. More specifically, we represent videos using sentences generated via video captioning methods and classes using sentences extracted from documents acquired through search engines on the Internet. Using these representations, we build a shared semantic space employing BERT-based embedders pre-trained in the paraphrasing task on multiple text datasets. The projection of both visual and semantic information onto this space is straightforward, as they are sentences, enabling classification using the nearest neighbor rule. We demonstrate that representing videos and labels with sentences alleviates the domain adaptation problem. Additionally, we show that word vectors are unsuitable for building the semantic embedding space of our descriptions. Our method outperforms the state-of-the-art performance on the UCF101 dataset by 3.3 p.p. in accuracy under the TruZe protocol and achieves competitive results on both the UCF101 and HMDB51 datasets under the conventional protocol (0/50\% - training/testing split). Our code is available at https://github.com/valterlej/zsarcap.

We are concerned with a challenging scenario in unpaired multiview video learning. In this case, the model aims to learn comprehensive multiview representations while the cross-view semantic information exhibits variations. We propose Semantics-based Unpaired Multiview Learning (SUM-L) to tackle this unpaired multiview learning problem. The key idea is to build cross-view pseudo-pairs and do view-invariant alignment by leveraging the semantic information of videos. To facilitate the data efficiency of multiview learning, we further perform video-text alignment for first-person and third-person videos, to fully leverage the semantic knowledge to improve video representations. Extensive experiments on multiple benchmark datasets verify the effectiveness of our framework. Our method also outperforms multiple existing view-alignment methods, under the more challenging scenario than typical paired or unpaired multimodal or multiview learning. Our code is available at https://github.com/wqtwjt1996/SUM-L.

Recent text-to-video generation approaches rely on computationally heavy training and require large-scale video datasets. In this paper, we introduce a new task of zero-shot text-to-video generation and propose a low-cost approach (without any training or optimization) by leveraging the power of existing text-to-image synthesis methods (e.g., Stable Diffusion), making them suitable for the video domain. Our key modifications include (i) enriching the latent codes of the generated frames with motion dynamics to keep the global scene and the background time consistent; and (ii) reprogramming frame-level self-attention using a new cross-frame attention of each frame on the first frame, to preserve the context, appearance, and identity of the foreground object. Experiments show that this leads to low overhead, yet high-quality and remarkably consistent video generation. Moreover, our approach is not limited to text-to-video synthesis but is also applicable to other tasks such as conditional and content-specialized video generation, and Video Instruct-Pix2Pix, i.e., instruction-guided video editing. As experiments show, our method performs comparably or sometimes better than recent approaches, despite not being trained on additional video data. Our code will be open sourced at: https://github.com/Picsart-AI-Research/Text2Video-Zero .

Pairwise pose estimation from images with little or no overlap is an open challenge in computer vision. Existing methods, even those trained on large-scale datasets, struggle in these scenarios due to the lack of identifiable correspondences or visual overlap. Inspired by the human ability to infer spatial relationships from diverse scenes, we propose a novel approach, InterPose, that leverages the rich priors encoded within pre-trained generative video models. We propose to use a video model to hallucinate intermediate frames between two input images, effectively creating a dense, visual transition, which significantly simplifies the problem of pose estimation. Since current video models can still produce implausible motion or inconsistent geometry, we introduce a self-consistency score that evaluates the consistency of pose predictions from sampled videos. We demonstrate that our approach generalizes among three state-of-the-art video models and show consistent improvements over the state-of-the-art DUSt3R on four diverse datasets encompassing indoor, outdoor, and object-centric scenes. Our findings suggest a promising avenue for improving pose estimation models by leveraging large generative models trained on vast amounts of video data, which is more readily available than 3D data. See our project page for results: https://inter-pose.github.io/.

Recently, breakthroughs in video modeling have allowed for controllable camera trajectories in generated videos. However, these methods cannot be directly applied to user-provided videos that are not generated by a video model. In this paper, we present ReCapture, a method for generating new videos with novel camera trajectories from a single user-provided video. Our method allows us to re-generate the reference video, with all its existing scene motion, from vastly different angles and with cinematic camera motion. Notably, using our method we can also plausibly hallucinate parts of the scene that were not observable in the reference video. Our method works by (1) generating a noisy anchor video with a new camera trajectory using multiview diffusion models or depth-based point cloud rendering and then (2) regenerating the anchor video into a clean and temporally consistent reangled video using our proposed masked video fine-tuning technique.

Contrastive language-image pretraining has shown great success in learning visual-textual joint representation from web-scale data, demonstrating remarkable "zero-shot" generalization ability for various image tasks. However, how to effectively expand such new language-image pretraining methods to video domains is still an open problem. In this work, we present a simple yet effective approach that adapts the pretrained language-image models to video recognition directly, instead of pretraining a new model from scratch. More concretely, to capture the long-range dependencies of frames along the temporal dimension, we propose a cross-frame attention mechanism that explicitly exchanges information across frames. Such module is lightweight and can be plugged into pretrained language-image models seamlessly. Moreover, we propose a video-specific prompting scheme, which leverages video content information for generating discriminative textual prompts. Extensive experiments demonstrate that our approach is effective and can be generalized to different video recognition scenarios. In particular, under fully-supervised settings, our approach achieves a top-1 accuracy of 87.1% on Kinectics-400, while using 12 times fewer FLOPs compared with Swin-L and ViViT-H. In zero-shot experiments, our approach surpasses the current state-of-the-art methods by +7.6% and +14.9% in terms of top-1 accuracy under two popular protocols. In few-shot scenarios, our approach outperforms previous best methods by +32.1% and +23.1% when the labeled data is extremely limited. Code and models are available at https://aka.ms/X-CLIP

We present a method for augmenting real-world videos with newly generated dynamic content. Given an input video and a simple user-provided text instruction describing the desired content, our method synthesizes dynamic objects or complex scene effects that naturally interact with the existing scene over time. The position, appearance, and motion of the new content are seamlessly integrated into the original footage while accounting for camera motion, occlusions, and interactions with other dynamic objects in the scene, resulting in a cohesive and realistic output video. We achieve this via a zero-shot, training-free framework that harnesses a pre-trained text-to-video diffusion transformer to synthesize the new content and a pre-trained Vision Language Model to envision the augmented scene in detail. Specifically, we introduce a novel inference-based method that manipulates features within the attention mechanism, enabling accurate localization and seamless integration of the new content while preserving the integrity of the original scene. Our method is fully automated, requiring only a simple user instruction. We demonstrate its effectiveness on a wide range of edits applied to real-world videos, encompassing diverse objects and scenarios involving both camera and object motion.

Building on the advances of language models, Large Multimodal Models (LMMs) have contributed significant improvements in video understanding. While the current video LMMs utilize advanced Large Language Models (LLMs), they rely on either image or video encoders to process visual inputs, each of which has its own limitations. Image encoders excel at capturing rich spatial details from frame sequences but lack explicit temporal context, which can be important in videos with intricate action sequences. On the other hand, video encoders provide temporal context but are often limited by computational constraints that lead to processing only sparse frames at lower resolutions, resulting in reduced contextual and spatial understanding. To this end, we introduce VideoGPT+, which combines the complementary benefits of the image encoder (for detailed spatial understanding) and the video encoder (for global temporal context modeling). The model processes videos by dividing them into smaller segments and applies an adaptive pooling strategy on features extracted by both image and video encoders. Our architecture showcases improved performance across multiple video benchmarks, including VCGBench, MVBench and Zero-shot question-answering. Further, we develop 112K video-instruction set using a novel semi-automatic annotation pipeline which further improves the model performance. Additionally, to comprehensively evaluate video LMMs, we present VCGBench-Diverse, covering 18 broad video categories such as lifestyle, sports, science, gaming, and surveillance videos. This benchmark with 4,354 question-answer pairs evaluates the generalization of existing LMMs on dense video captioning, spatial and temporal understanding, and complex reasoning, ensuring comprehensive assessment across diverse video types and dynamics. Code: https://github.com/mbzuai-oryx/VideoGPT-plus.

We introduce Ego4D, a massive-scale egocentric video dataset and benchmark suite. It offers 3,670 hours of daily-life activity video spanning hundreds of scenarios (household, outdoor, workplace, leisure, etc.) captured by 931 unique camera wearers from 74 worldwide locations and 9 different countries. The approach to collection is designed to uphold rigorous privacy and ethics standards with consenting participants and robust de-identification procedures where relevant. Ego4D dramatically expands the volume of diverse egocentric video footage publicly available to the research community. Portions of the video are accompanied by audio, 3D meshes of the environment, eye gaze, stereo, and/or synchronized videos from multiple egocentric cameras at the same event. Furthermore, we present a host of new benchmark challenges centered around understanding the first-person visual experience in the past (querying an episodic memory), present (analyzing hand-object manipulation, audio-visual conversation, and social interactions), and future (forecasting activities). By publicly sharing this massive annotated dataset and benchmark suite, we aim to push the frontier of first-person perception. Project page: https://ego4d-data.org/

We introduce VideoPrism, a general-purpose video encoder that tackles diverse video understanding tasks with a single frozen model. We pretrain VideoPrism on a heterogeneous corpus containing 36M high-quality video-caption pairs and 582M video clips with noisy parallel text (e.g., ASR transcripts). The pretraining approach improves upon masked autoencoding by global-local distillation of semantic video embeddings and a token shuffling scheme, enabling VideoPrism to focus primarily on the video modality while leveraging the invaluable text associated with videos. We extensively test VideoPrism on four broad groups of video understanding tasks, from web video question answering to CV for science, achieving state-of-the-art performance on 30 out of 33 video understanding benchmarks.

We address the problem of video representation learning without human-annotated labels. While previous efforts address the problem by designing novel self-supervised tasks using video data, the learned features are merely on a frame-by-frame basis, which are not applicable to many video analytic tasks where spatio-temporal features are prevailing. In this paper we propose a novel self-supervised approach to learn spatio-temporal features for video representation. Inspired by the success of two-stream approaches in video classification, we propose to learn visual features by regressing both motion and appearance statistics along spatial and temporal dimensions, given only the input video data. Specifically, we extract statistical concepts (fast-motion region and the corresponding dominant direction, spatio-temporal color diversity, dominant color, etc.) from simple patterns in both spatial and temporal domains. Unlike prior puzzles that are even hard for humans to solve, the proposed approach is consistent with human inherent visual habits and therefore easy to answer. We conduct extensive experiments with C3D to validate the effectiveness of our proposed approach. The experiments show that our approach can significantly improve the performance of C3D when applied to video classification tasks. Code is available at https://github.com/laura-wang/video_repres_mas.

Stimulated by the sophisticated reasoning capabilities of recent Large Language Models (LLMs), a variety of strategies for bridging video modality have been devised. A prominent strategy involves Video Language Models (VideoLMs), which train a learnable interface with video data to connect advanced vision encoders with LLMs. Recently, an alternative strategy has surfaced, employing readily available foundation models, such as VideoLMs and LLMs, across multiple stages for modality bridging. In this study, we introduce a simple yet novel strategy where only a single Vision Language Model (VLM) is utilized. Our starting point is the plain insight that a video comprises a series of images, or frames, interwoven with temporal information. The essence of video comprehension lies in adeptly managing the temporal aspects along with the spatial details of each frame. Initially, we transform a video into a single composite image by arranging multiple frames in a grid layout. The resulting single image is termed as an image grid. This format, while maintaining the appearance of a solitary image, effectively retains temporal information within the grid structure. Therefore, the image grid approach enables direct application of a single high-performance VLM without necessitating any video-data training. Our extensive experimental analysis across ten zero-shot video question answering benchmarks, including five open-ended and five multiple-choice benchmarks, reveals that the proposed Image Grid Vision Language Model (IG-VLM) surpasses the existing methods in nine out of ten benchmarks.

In text-video retrieval, the objective is to learn a cross-modal similarity function between a text and a video that ranks relevant text-video pairs higher than irrelevant pairs. However, videos inherently express a much wider gamut of information than texts. Instead, texts often capture sub-regions of entire videos and are most semantically similar to certain frames within videos. Therefore, for a given text, a retrieval model should focus on the text's most semantically similar video sub-regions to make a more relevant comparison. Yet, most existing works aggregate entire videos without directly considering text. Common text-agnostic aggregations schemes include mean-pooling or self-attention over the frames, but these are likely to encode misleading visual information not described in the given text. To address this, we propose a cross-modal attention model called X-Pool that reasons between a text and the frames of a video. Our core mechanism is a scaled dot product attention for a text to attend to its most semantically similar frames. We then generate an aggregated video representation conditioned on the text's attention weights over the frames. We evaluate our method on three benchmark datasets of MSR-VTT, MSVD and LSMDC, achieving new state-of-the-art results by up to 12% in relative improvement in Recall@1. Our findings thereby highlight the importance of joint text-video reasoning to extract important visual cues according to text. Full code and demo can be found at: https://layer6ai-labs.github.io/xpool/

Vision-language large models have achieved remarkable success in various multi-modal tasks, yet applying them to video understanding remains challenging due to the inherent complexity and computational demands of video data. While training-based video-LLMs deliver high performance, they often require substantial resources for training and inference. Conversely, training-free approaches offer a more efficient alternative by adapting pre-trained image-LLMs models for video tasks without additional training, but they face inference efficiency bottlenecks due to the large number of visual tokens generated from video frames. In this work, we present a novel prompt-guided visual perception framework (abbreviated as Free Video-LLM) for efficient inference of training-free video LLMs. The proposed framework decouples spatial-temporal dimension and performs temporal frame sampling and spatial RoI cropping respectively based on task-specific prompts. Our method effectively reduces the number of visual tokens while maintaining high performance across multiple video question-answering benchmarks. Extensive experiments demonstrate that our approach achieves competitive results with significantly fewer tokens, offering an optimal trade-off between accuracy and computational efficiency compared to state-of-the-art video LLMs. The code will be available at https://github.com/contrastive/FreeVideoLLM.

Tracking human object interaction from videos is important to understand human behavior from the rapidly growing stream of video data. Previous video-based methods require predefined object templates while single-image-based methods are template-free but lack temporal consistency. In this paper, we present a method to track human object interaction without any object shape templates. We decompose the 4D tracking problem into per-frame pose tracking and canonical shape optimization. We first apply a single-view reconstruction method to obtain temporally-inconsistent per-frame interaction reconstructions. Then, for the human, we propose an efficient autoencoder to predict SMPL vertices directly from the per-frame reconstructions, introducing temporally consistent correspondence. For the object, we introduce a pose estimator that leverages temporal information to predict smooth object rotations under occlusions. To train our model, we propose a method to generate synthetic interaction videos and synthesize in total 10 hour videos of 8.5k sequences with full 3D ground truth. Experiments on BEHAVE and InterCap show that our method significantly outperforms previous template-based video tracking and single-frame reconstruction methods. Our proposed synthetic video dataset also allows training video-based methods that generalize to real-world videos. Our code and dataset will be publicly released.

We present CAT4D, a method for creating 4D (dynamic 3D) scenes from monocular video. CAT4D leverages a multi-view video diffusion model trained on a diverse combination of datasets to enable novel view synthesis at any specified camera poses and timestamps. Combined with a novel sampling approach, this model can transform a single monocular video into a multi-view video, enabling robust 4D reconstruction via optimization of a deformable 3D Gaussian representation. We demonstrate competitive performance on novel view synthesis and dynamic scene reconstruction benchmarks, and highlight the creative capabilities for 4D scene generation from real or generated videos. See our project page for results and interactive demos: cat-4d.github.io.

Despite constituting 65% of all internet traffic in 2023, video content is underrepresented in generative AI research. Meanwhile, recent large language models (LLMs) have become increasingly integrated with capabilities in the visual modality. Integrating video with LLMs is a natural next step, so how can this gap be bridged? To advance video reasoning, we propose a new research direction of VideoCOT on video keyframes, which leverages the multimodal generative abilities of vision-language models to enhance video reasoning while reducing the computational complexity of processing hundreds or thousands of frames. We introduce VIP, an inference-time dataset that can be used to evaluate VideoCOT, containing 1) a variety of real-life videos with keyframes and corresponding unstructured and structured scene descriptions, and 2) two new video reasoning tasks: video infilling and scene prediction. We benchmark various vision-language models on VIP, demonstrating the potential to use vision-language models and LLMs to enhance video chain of thought reasoning.

We present xGen-MM-Vid (BLIP-3-Video): a multimodal language model for videos, particularly designed to efficiently capture temporal information over multiple frames. BLIP-3-Video takes advantage of the 'temporal encoder' in addition to the conventional visual tokenizer, which maps a sequence of tokens over multiple frames into a compact set of visual tokens. This enables BLIP3-Video to use much fewer visual tokens than its competing models (e.g., 32 vs. 4608 tokens). We explore different types of temporal encoders, including learnable spatio-temporal pooling as well as sequential models like Token Turing Machines. We experimentally confirm that BLIP-3-Video obtains video question-answering accuracies comparable to much larger state-of-the-art models (e.g., 34B), while being much smaller (i.e., 4B) and more efficient by using fewer visual tokens. The project website is at https://www.salesforceairesearch.com/opensource/xGen-MM-Vid/index.html

Self-supervised monocular depth estimation networks are trained to predict scene depth using nearby frames as a supervision signal during training. However, for many applications, sequence information in the form of video frames is also available at test time. The vast majority of monocular networks do not make use of this extra signal, thus ignoring valuable information that could be used to improve the predicted depth. Those that do, either use computationally expensive test-time refinement techniques or off-the-shelf recurrent networks, which only indirectly make use of the geometric information that is inherently available. We propose ManyDepth, an adaptive approach to dense depth estimation that can make use of sequence information at test time, when it is available. Taking inspiration from multi-view stereo, we propose a deep end-to-end cost volume based approach that is trained using self-supervision only. We present a novel consistency loss that encourages the network to ignore the cost volume when it is deemed unreliable, e.g. in the case of moving objects, and an augmentation scheme to cope with static cameras. Our detailed experiments on both KITTI and Cityscapes show that we outperform all published self-supervised baselines, including those that use single or multiple frames at test time.

Recently, video generation has attracted massive attention and yielded noticeable outcomes. Concerning the characteristics of video, multi-text conditioning incorporating sequential events is necessary for next-step video generation. In this work, we propose a novel multi-text video generation~(MTVG) by directly utilizing a pre-trained diffusion-based text-to-video~(T2V) generation model without additional fine-tuning. To generate consecutive video segments, visual consistency generated by distinct prompts is necessary with diverse variations, such as motion and content-related transitions. Our proposed MTVG includes Dynamic Noise and Last Frame Aware Inversion which reinitialize the noise latent to preserve visual coherence between videos of different prompts and prevent repetitive motion or contents. Furthermore, we present Structure Guiding Sampling to maintain the global appearance across the frames in a single video clip, where we leverage iterative latent updates across the preceding frame. Additionally, our Prompt Generator allows for arbitrary format of text conditions consisting of diverse events. As a result, our extensive experiments, including diverse transitions of descriptions, demonstrate that our proposed methods show superior generated outputs in terms of semantically coherent and temporally seamless video.Video examples are available in our project page: https://kuai-lab.github.io/mtvg-page.

Video view synthesis, allowing for the creation of visually appealing frames from arbitrary viewpoints and times, offers immersive viewing experiences. Neural radiance fields, particularly NeRF, initially developed for static scenes, have spurred the creation of various methods for video view synthesis. However, the challenge for video view synthesis arises from motion blur, a consequence of object or camera movement during exposure, which hinders the precise synthesis of sharp spatio-temporal views. In response, we propose a novel dynamic deblurring NeRF framework for blurry monocular video, called DyBluRF, consisting of an Interleave Ray Refinement (IRR) stage and a Motion Decomposition-based Deblurring (MDD) stage. Our DyBluRF is the first that addresses and handles the novel view synthesis for blurry monocular video. The IRR stage jointly reconstructs dynamic 3D scenes and refines the inaccurate camera pose information to combat imprecise pose information extracted from the given blurry frames. The MDD stage is a novel incremental latent sharp-rays prediction (ILSP) approach for the blurry monocular video frames by decomposing the latent sharp rays into global camera motion and local object motion components. Extensive experimental results demonstrate that our DyBluRF outperforms qualitatively and quantitatively the very recent state-of-the-art methods. Our project page including source codes and pretrained model are publicly available at https://kaist-viclab.github.io/dyblurf-site/.

4D video control is essential in video generation as it enables the use of sophisticated lens techniques, such as multi-camera shooting and dolly zoom, which are currently unsupported by existing methods. Training a video Diffusion Transformer (DiT) directly to control 4D content requires expensive multi-view videos. Inspired by Monocular Dynamic novel View Synthesis (MDVS) that optimizes a 4D representation and renders videos according to different 4D elements, such as camera pose and object motion editing, we bring pseudo 4D Gaussian fields to video generation. Specifically, we propose a novel framework that constructs a pseudo 4D Gaussian field with dense 3D point tracking and renders the Gaussian field for all video frames. Then we finetune a pretrained DiT to generate videos following the guidance of the rendered video, dubbed as GS-DiT. To boost the training of the GS-DiT, we also propose an efficient Dense 3D Point Tracking (D3D-PT) method for the pseudo 4D Gaussian field construction. Our D3D-PT outperforms SpatialTracker, the state-of-the-art sparse 3D point tracking method, in accuracy and accelerates the inference speed by two orders of magnitude. During the inference stage, GS-DiT can generate videos with the same dynamic content while adhering to different camera parameters, addressing a significant limitation of current video generation models. GS-DiT demonstrates strong generalization capabilities and extends the 4D controllability of Gaussian splatting to video generation beyond just camera poses. It supports advanced cinematic effects through the manipulation of the Gaussian field and camera intrinsics, making it a powerful tool for creative video production. Demos are available at https://wkbian.github.io/Projects/GS-DiT/.

Monocular 3D object detection (Mono3D) holds noteworthy promise for autonomous driving applications owing to the cost-effectiveness and rich visual context of monocular camera sensors. However, depth ambiguity poses a significant challenge, as it requires extracting precise 3D scene geometry from a single image, resulting in suboptimal performance when transferring knowledge from a LiDAR-based teacher model to a camera-based student model. To address this issue, we introduce {\em Monocular Teaching Assistant Knowledge Distillation (MonoTAKD)} to enhance 3D perception in Mono3D. Our approach presents a robust camera-based teaching assistant model that effectively bridges the representation gap between different modalities for teacher and student models, addressing the challenge of inaccurate depth estimation. By defining 3D spatial cues as residual features that capture the differences between the teacher and the teaching assistant models, we leverage these cues into the student model, improving its 3D perception capabilities. Experimental results show that our MonoTAKD achieves state-of-the-art performance on the KITTI3D dataset. Additionally, we evaluate the performance on nuScenes and KITTI raw datasets to demonstrate the generalization of our model to multi-view 3D and unsupervised data settings. Our code will be available at https://github.com/hoiliu-0801/MonoTAKD.

Contrastive language-image pretraining (CLIP) has demonstrated remarkable success in various image tasks. However, how to extend CLIP with effective temporal modeling is still an open and crucial problem. Existing factorized or joint spatial-temporal modeling trades off between the efficiency and performance. While modeling temporal information within straight through tube is widely adopted in literature, we find that simple frame alignment already provides enough essence without temporal attention. To this end, in this paper, we proposed a novel Implicit Learnable Alignment (ILA) method, which minimizes the temporal modeling effort while achieving incredibly high performance. Specifically, for a frame pair, an interactive point is predicted in each frame, serving as a mutual information rich region. By enhancing the features around the interactive point, two frames are implicitly aligned. The aligned features are then pooled into a single token, which is leveraged in the subsequent spatial self-attention. Our method allows eliminating the costly or insufficient temporal self-attention in video. Extensive experiments on benchmarks demonstrate the superiority and generality of our module. Particularly, the proposed ILA achieves a top-1 accuracy of 88.7% on Kinetics-400 with much fewer FLOPs compared with Swin-L and ViViT-H. Code is released at https://github.com/Francis-Rings/ILA .

We focus on the task of Automatic Live Video Commenting (ALVC), which aims to generate real-time video comments with both video frames and other viewers' comments as inputs. A major challenge in this task is how to properly leverage the rich and diverse information carried by video and text. In this paper, we aim to collect diversified information from video and text for informative comment generation. To achieve this, we propose a Diversified Co-Attention (DCA) model for this task. Our model builds bidirectional interactions between video frames and surrounding comments from multiple perspectives via metric learning, to collect a diversified and informative context for comment generation. We also propose an effective parameter orthogonalization technique to avoid excessive overlap of information learned from different perspectives. Results show that our approach outperforms existing methods in the ALVC task, achieving new state-of-the-art results.

Building video-language foundation models is costly and difficult due to the redundant nature of video data and the lack of high-quality video-language datasets. In this paper, we propose an efficient framework to harvest video foundation models from image ones. Our method is intuitively simple by randomly dropping input video patches and masking out input text during the post-pretraining procedure. The patch dropping boosts the training efficiency significantly and text masking enforces the learning of cross-modal fusion. We conduct extensive experiments to validate the effectiveness of our method on a wide range of video-language downstream tasks including various zero-shot tasks, video question answering, and video-text retrieval. Despite its simplicity, our method achieves state-of-the-art performances, which are comparable to some heavily pretrained video foundation models. Our method is extremely efficient and can be trained in less than one day on 8 GPUs, requiring only WebVid-10M as pretraining data. We hope our method can serve as a simple yet strong counterpart for prevalent video foundation models, provide useful insights when building them, and make large pretrained models more accessible and sustainable. This is part of the InternVideo project https://github.com/OpenGVLab/InternVideo.

The task of assigning semantic classes and track identities to every pixel in a video is called video panoptic segmentation. Our work is the first that targets this task in a real-world setting requiring dense interpretation in both spatial and temporal domains. As the ground-truth for this task is difficult and expensive to obtain, existing datasets are either constructed synthetically or only sparsely annotated within short video clips. To overcome this, we introduce a new benchmark encompassing two datasets, KITTI-STEP, and MOTChallenge-STEP. The datasets contain long video sequences, providing challenging examples and a test-bed for studying long-term pixel-precise segmentation and tracking under real-world conditions. We further propose a novel evaluation metric Segmentation and Tracking Quality (STQ) that fairly balances semantic and tracking aspects of this task and is more appropriate for evaluating sequences of arbitrary length. Finally, we provide several baselines to evaluate the status of existing methods on this new challenging dataset. We have made our datasets, metric, benchmark servers, and baselines publicly available, and hope this will inspire future research.

We introduce the MAsked Generative VIdeo Transformer, MAGVIT, to tackle various video synthesis tasks with a single model. We introduce a 3D tokenizer to quantize a video into spatial-temporal visual tokens and propose an embedding method for masked video token modeling to facilitate multi-task learning. We conduct extensive experiments to demonstrate the quality, efficiency, and flexibility of MAGVIT. Our experiments show that (i) MAGVIT performs favorably against state-of-the-art approaches and establishes the best-published FVD on three video generation benchmarks, including the challenging Kinetics-600. (ii) MAGVIT outperforms existing methods in inference time by two orders of magnitude against diffusion models and by 60x against autoregressive models. (iii) A single MAGVIT model supports ten diverse generation tasks and generalizes across videos from different visual domains. The source code and trained models will be released to the public at https://magvit.cs.cmu.edu.

Synthesizing motion-rich and temporally consistent videos remains a challenge in artificial intelligence, especially when dealing with extended durations. Existing text-to-video (T2V) models commonly employ spatial cross-attention for text control, equivalently guiding different frame generations without frame-specific textual guidance. Thus, the model's capacity to comprehend the temporal logic conveyed in prompts and generate videos with coherent motion is restricted. To tackle this limitation, we introduce FancyVideo, an innovative video generator that improves the existing text-control mechanism with the well-designed Cross-frame Textual Guidance Module (CTGM). Specifically, CTGM incorporates the Temporal Information Injector (TII), Temporal Affinity Refiner (TAR), and Temporal Feature Booster (TFB) at the beginning, middle, and end of cross-attention, respectively, to achieve frame-specific textual guidance. Firstly, TII injects frame-specific information from latent features into text conditions, thereby obtaining cross-frame textual conditions. Then, TAR refines the correlation matrix between cross-frame textual conditions and latent features along the time dimension. Lastly, TFB boosts the temporal consistency of latent features. Extensive experiments comprising both quantitative and qualitative evaluations demonstrate the effectiveness of FancyVideo. Our approach achieves state-of-the-art T2V generation results on the EvalCrafter benchmark and facilitates the synthesis of dynamic and consistent videos. The video show results can be available at https://fancyvideo.github.io/, and we will make our code and model weights publicly available.

The core of video understanding tasks, such as recognition, captioning, and tracking, is to automatically detect objects or actions in a video and analyze their temporal evolution. Despite sharing a common goal, different tasks often rely on distinct model architectures and annotation formats. In contrast, natural language processing benefits from a unified output space, i.e., text sequences, which simplifies the training of powerful foundational language models, such as GPT-3, with extensive training corpora. Inspired by this, we seek to unify the output space of video understanding tasks by using languages as labels and additionally introducing time and box tokens. In this way, a variety of video tasks could be formulated as video-grounded token generation. This enables us to address various types of video tasks, including classification (such as action recognition), captioning (covering clip captioning, video question answering, and dense video captioning), and localization tasks (such as visual object tracking) within a fully shared encoder-decoder architecture, following a generative framework. Through comprehensive experiments, we demonstrate such a simple and straightforward idea is quite effective and can achieve state-of-the-art or competitive results on seven video benchmarks, providing a novel perspective for more universal video understanding. Code is available at https://github.com/wangjk666/OmniVid.

Video super-resolution (VSR) aims to restore a sequence of high-resolution (HR) frames from their low-resolution (LR) counterparts. Although some progress has been made, there are grand challenges to effectively utilize temporal dependency in entire video sequences. Existing approaches usually align and aggregate video frames from limited adjacent frames (e.g., 5 or 7 frames), which prevents these approaches from satisfactory results. In this paper, we take one step further to enable effective spatio-temporal learning in videos. We propose a novel Trajectory-aware Transformer for Video Super-Resolution (TTVSR). In particular, we formulate video frames into several pre-aligned trajectories which consist of continuous visual tokens. For a query token, self-attention is only learned on relevant visual tokens along spatio-temporal trajectories. Compared with vanilla vision Transformers, such a design significantly reduces the computational cost and enables Transformers to model long-range features. We further propose a cross-scale feature tokenization module to overcome scale-changing problems that often occur in long-range videos. Experimental results demonstrate the superiority of the proposed TTVSR over state-of-the-art models, by extensive quantitative and qualitative evaluations in four widely-used video super-resolution benchmarks. Both code and pre-trained models can be downloaded at https://github.com/researchmm/TTVSR.

We present an unsupervised learning framework for the task of monocular depth and camera motion estimation from unstructured video sequences. We achieve this by simultaneously training depth and camera pose estimation networks using the task of view synthesis as the supervisory signal. The networks are thus coupled via the view synthesis objective during training, but can be applied independently at test time. Empirical evaluation on the KITTI dataset demonstrates the effectiveness of our approach: 1) monocular depth performing comparably with supervised methods that use either ground-truth pose or depth for training, and 2) pose estimation performing favorably with established SLAM systems under comparable input settings.

Recent advancements in video large multimodal models (LMMs) have significantly improved their video understanding and reasoning capabilities. However, their performance drops on out-of-distribution (OOD) tasks that are underrepresented in training data. Traditional methods like fine-tuning on OOD datasets are impractical due to high computational costs. While In-context learning (ICL) with demonstration examples has shown promising generalization performance in language tasks and image-language tasks without fine-tuning, applying ICL to video-language tasks faces challenges due to the limited context length in Video LMMs, as videos require longer token lengths. To address these issues, we propose VideoICL, a novel video in-context learning framework for OOD tasks that introduces a similarity-based relevant example selection strategy and a confidence-based iterative inference approach. This allows to select the most relevant examples and rank them based on similarity, to be used for inference. If the generated response has low confidence, our framework selects new examples and performs inference again, iteratively refining the results until a high-confidence response is obtained. This approach improves OOD video understanding performance by extending effective context length without incurring high costs. The experimental results on multiple benchmarks demonstrate significant performance gains, especially in domain-specific scenarios, laying the groundwork for broader video comprehension applications. Code will be released at https://github.com/KangsanKim07/VideoICL

In the quest for artificial general intelligence, Multi-modal Large Language Models (MLLMs) have emerged as a focal point in recent advancements. However, the predominant focus remains on developing their capabilities in static image understanding. The potential of MLLMs in processing sequential visual data is still insufficiently explored, highlighting the absence of a comprehensive, high-quality assessment of their performance. In this paper, we introduce Video-MME, the first-ever full-spectrum, Multi-Modal Evaluation benchmark of MLLMs in Video analysis. Our work distinguishes from existing benchmarks through four key features: 1) Diversity in video types, spanning 6 primary visual domains with 30 subfields to ensure broad scenario generalizability; 2) Duration in temporal dimension, encompassing both short-, medium-, and long-term videos, ranging from 11 seconds to 1 hour, for robust contextual dynamics; 3) Breadth in data modalities, integrating multi-modal inputs besides video frames, including subtitles and audios, to unveil the all-round capabilities of MLLMs; 4) Quality in annotations, utilizing rigorous manual labeling by expert annotators to facilitate precise and reliable model assessment. 900 videos with a total of 256 hours are manually selected and annotated by repeatedly viewing all the video content, resulting in 2,700 question-answer pairs. With Video-MME, we extensively evaluate various state-of-the-art MLLMs, including GPT-4 series and Gemini 1.5 Pro, as well as open-source image models like InternVL-Chat-V1.5 and video models like LLaVA-NeXT-Video. Our experiments reveal that Gemini 1.5 Pro is the best-performing commercial model, significantly outperforming the open-source models. Our dataset along with these findings underscores the need for further improvements in handling longer sequences and multi-modal data. Project Page: https://video-mme.github.io

Despite tremendous recent progress, generative video models still struggle to capture real-world motion, dynamics, and physics. We show that this limitation arises from the conventional pixel reconstruction objective, which biases models toward appearance fidelity at the expense of motion coherence. To address this, we introduce VideoJAM, a novel framework that instills an effective motion prior to video generators, by encouraging the model to learn a joint appearance-motion representation. VideoJAM is composed of two complementary units. During training, we extend the objective to predict both the generated pixels and their corresponding motion from a single learned representation. During inference, we introduce Inner-Guidance, a mechanism that steers the generation toward coherent motion by leveraging the model's own evolving motion prediction as a dynamic guidance signal. Notably, our framework can be applied to any video model with minimal adaptations, requiring no modifications to the training data or scaling of the model. VideoJAM achieves state-of-the-art performance in motion coherence, surpassing highly competitive proprietary models while also enhancing the perceived visual quality of the generations. These findings emphasize that appearance and motion can be complementary and, when effectively integrated, enhance both the visual quality and the coherence of video generation. Project website: https://hila-chefer.github.io/videojam-paper.github.io/

While large-scale image-text pretrained models such as CLIP have been used for multiple video-level tasks on trimmed videos, their use for temporal localization in untrimmed videos is still a relatively unexplored task. We design a new approach for this called UnLoc, which uses pretrained image and text towers, and feeds tokens to a video-text fusion model. The output of the fusion module are then used to construct a feature pyramid in which each level connects to a head to predict a per-frame relevancy score and start/end time displacements. Unlike previous works, our architecture enables Moment Retrieval, Temporal Localization, and Action Segmentation with a single stage model, without the need for action proposals, motion based pretrained features or representation masking. Unlike specialized models, we achieve state of the art results on all three different localization tasks with a unified approach. Code will be available at: https://github.com/google-research/scenic.

We introduce Presto, a novel video diffusion model designed to generate 15-second videos with long-range coherence and rich content. Extending video generation methods to maintain scenario diversity over long durations presents significant challenges. To address this, we propose a Segmented Cross-Attention (SCA) strategy, which splits hidden states into segments along the temporal dimension, allowing each segment to cross-attend to a corresponding sub-caption. SCA requires no additional parameters, enabling seamless incorporation into current DiT-based architectures. To facilitate high-quality long video generation, we build the LongTake-HD dataset, consisting of 261k content-rich videos with scenario coherence, annotated with an overall video caption and five progressive sub-captions. Experiments show that our Presto achieves 78.5% on the VBench Semantic Score and 100% on the Dynamic Degree, outperforming existing state-of-the-art video generation methods. This demonstrates that our proposed Presto significantly enhances content richness, maintains long-range coherence, and captures intricate textual details. More details are displayed on our project page: https://presto-video.github.io/.

We present Stable Video 4D (SV4D), a latent video diffusion model for multi-frame and multi-view consistent dynamic 3D content generation. Unlike previous methods that rely on separately trained generative models for video generation and novel view synthesis, we design a unified diffusion model to generate novel view videos of dynamic 3D objects. Specifically, given a monocular reference video, SV4D generates novel views for each video frame that are temporally consistent. We then use the generated novel view videos to optimize an implicit 4D representation (dynamic NeRF) efficiently, without the need for cumbersome SDS-based optimization used in most prior works. To train our unified novel view video generation model, we curated a dynamic 3D object dataset from the existing Objaverse dataset. Extensive experimental results on multiple datasets and user studies demonstrate SV4D's state-of-the-art performance on novel-view video synthesis as well as 4D generation compared to prior works.

RGB video object tracking is a fundamental task in computer vision. Its effectiveness can be improved using depth information, particularly for handling motion-blurred target. However, depth information is often missing in commonly used tracking benchmarks. In this work, we propose a new framework that leverages monocular depth estimation to counter the challenges of tracking targets that are out of view or affected by motion blur in RGB video sequences. Specifically, our work introduces following contributions. To the best of our knowledge, we are the first to propose a depth attention mechanism and to formulate a simple framework that allows seamlessly integration of depth information with state of the art tracking algorithms, without RGB-D cameras, elevating accuracy and robustness. We provide extensive experiments on six challenging tracking benchmarks. Our results demonstrate that our approach provides consistent gains over several strong baselines and achieves new SOTA performance. We believe that our method will open up new possibilities for more sophisticated VOT solutions in real-world scenarios. Our code and models are publicly released: https://github.com/LiuYuML/Depth-Attention.

We capitalize on large amounts of unlabeled video in order to learn a model of scene dynamics for both video recognition tasks (e.g. action classification) and video generation tasks (e.g. future prediction). We propose a generative adversarial network for video with a spatio-temporal convolutional architecture that untangles the scene's foreground from the background. Experiments suggest this model can generate tiny videos up to a second at full frame rate better than simple baselines, and we show its utility at predicting plausible futures of static images. Moreover, experiments and visualizations show the model internally learns useful features for recognizing actions with minimal supervision, suggesting scene dynamics are a promising signal for representation learning. We believe generative video models can impact many applications in video understanding and simulation.

We present a unified controllable video generation approach AnimateAnything that facilitates precise and consistent video manipulation across various conditions, including camera trajectories, text prompts, and user motion annotations. Specifically, we carefully design a multi-scale control feature fusion network to construct a common motion representation for different conditions. It explicitly converts all control information into frame-by-frame optical flows. Then we incorporate the optical flows as motion priors to guide final video generation. In addition, to reduce the flickering issues caused by large-scale motion, we propose a frequency-based stabilization module. It can enhance temporal coherence by ensuring the video's frequency domain consistency. Experiments demonstrate that our method outperforms the state-of-the-art approaches. For more details and videos, please refer to the webpage: https://yu-shaonian.github.io/Animate_Anything/.

Recent advancements in video diffusion models have shown exceptional abilities in simulating real-world dynamics and maintaining 3D consistency. This progress inspires us to investigate the potential of these models to ensure dynamic consistency across various viewpoints, a highly desirable feature for applications such as virtual filming. Unlike existing methods focused on multi-view generation of single objects for 4D reconstruction, our interest lies in generating open-world videos from arbitrary viewpoints, incorporating 6 DoF camera poses. To achieve this, we propose a plug-and-play module that enhances a pre-trained text-to-video model for multi-camera video generation, ensuring consistent content across different viewpoints. Specifically, we introduce a multi-view synchronization module to maintain appearance and geometry consistency across these viewpoints. Given the scarcity of high-quality training data, we design a hybrid training scheme that leverages multi-camera images and monocular videos to supplement Unreal Engine-rendered multi-camera videos. Furthermore, our method enables intriguing extensions, such as re-rendering a video from novel viewpoints. We also release a multi-view synchronized video dataset, named SynCamVideo-Dataset. Project page: https://jianhongbai.github.io/SynCamMaster/.

We present LLaVA-OneVision, a family of open large multimodal models (LMMs) developed by consolidating our insights into data, models, and visual representations in the LLaVA-NeXT blog series. Our experimental results demonstrate that LLaVA-OneVision is the first single model that can simultaneously push the performance boundaries of open LMMs in three important computer vision scenarios: single-image, multi-image, and video scenarios. Importantly, the design of LLaVA-OneVision allows strong transfer learning across different modalities/scenarios, yielding new emerging capabilities. In particular, strong video understanding and cross-scenario capabilities are demonstrated through task transfer from images to videos.

A well-known dilemma in large vision-language models (e.g., GPT-4, LLaVA) is that while increasing the number of vision tokens generally enhances visual understanding, it also significantly raises memory and computational costs, especially in long-term, dense video frame streaming scenarios. Although learnable approaches like Q-Former and Perceiver Resampler have been developed to reduce the vision token burden, they overlook the context causally modeled by LLMs (i.e., key-value cache), potentially leading to missed visual cues when addressing user queries. In this paper, we introduce a novel approach to reduce vision compute by leveraging redundant vision tokens "skipping layers" rather than decreasing the number of vision tokens. Our method, VideoLLM-MoD, is inspired by mixture-of-depths LLMs and addresses the challenge of numerous vision tokens in long-term or streaming video. Specifically, for each transformer layer, we learn to skip the computation for a high proportion (e.g., 80\%) of vision tokens, passing them directly to the next layer. This approach significantly enhances model efficiency, achieving approximately \textasciitilde42\% time and \textasciitilde30\% memory savings for the entire training. Moreover, our method reduces the computation in the context and avoid decreasing the vision tokens, thus preserving or even improving performance compared to the vanilla model. We conduct extensive experiments to demonstrate the effectiveness of VideoLLM-MoD, showing its state-of-the-art results on multiple benchmarks, including narration, forecasting, and summarization tasks in COIN, Ego4D, and Ego-Exo4D datasets.

We present Phenaki, a model capable of realistic video synthesis, given a sequence of textual prompts. Generating videos from text is particularly challenging due to the computational cost, limited quantities of high quality text-video data and variable length of videos. To address these issues, we introduce a new model for learning video representation which compresses the video to a small representation of discrete tokens. This tokenizer uses causal attention in time, which allows it to work with variable-length videos. To generate video tokens from text we are using a bidirectional masked transformer conditioned on pre-computed text tokens. The generated video tokens are subsequently de-tokenized to create the actual video. To address data issues, we demonstrate how joint training on a large corpus of image-text pairs as well as a smaller number of video-text examples can result in generalization beyond what is available in the video datasets. Compared to the previous video generation methods, Phenaki can generate arbitrary long videos conditioned on a sequence of prompts (i.e. time variable text or a story) in open domain. To the best of our knowledge, this is the first time a paper studies generating videos from time variable prompts. In addition, compared to the per-frame baselines, the proposed video encoder-decoder computes fewer tokens per video but results in better spatio-temporal consistency.

Despite the recent progress in text-to-video generation, existing studies usually overlook the issue that only spatial contents but not temporal motions in synthesized videos are under the control of text. Towards such a challenge, this work presents a practical system, named LivePhoto, which allows users to animate an image of their interest with text descriptions. We first establish a strong baseline that helps a well-learned text-to-image generator (i.e., Stable Diffusion) take an image as a further input. We then equip the improved generator with a motion module for temporal modeling and propose a carefully designed training pipeline to better link texts and motions. In particular, considering the facts that (1) text can only describe motions roughly (e.g., regardless of the moving speed) and (2) text may include both content and motion descriptions, we introduce a motion intensity estimation module as well as a text re-weighting module to reduce the ambiguity of text-to-motion mapping. Empirical evidence suggests that our approach is capable of well decoding motion-related textual instructions into videos, such as actions, camera movements, or even conjuring new contents from thin air (e.g., pouring water into an empty glass). Interestingly, thanks to the proposed intensity learning mechanism, our system offers users an additional control signal (i.e., the motion intensity) besides text for video customization.

Recently, the community has made tremendous progress in developing effective methods for point cloud video understanding that learn from massive amounts of labeled data. However, annotating point cloud videos is usually notoriously expensive. Moreover, training via one or only a few traditional tasks (e.g., classification) may be insufficient to learn subtle details of the spatio-temporal structure existing in point cloud videos. In this paper, we propose a Masked Spatio-Temporal Structure Prediction (MaST-Pre) method to capture the structure of point cloud videos without human annotations. MaST-Pre is based on spatio-temporal point-tube masking and consists of two self-supervised learning tasks. First, by reconstructing masked point tubes, our method is able to capture the appearance information of point cloud videos. Second, to learn motion, we propose a temporal cardinality difference prediction task that estimates the change in the number of points within a point tube. In this way, MaST-Pre is forced to model the spatial and temporal structure in point cloud videos. Extensive experiments on MSRAction-3D, NTU-RGBD, NvGesture, and SHREC'17 demonstrate the effectiveness of the proposed method.

In recent years, many video tasks have achieved breakthroughs by utilizing the vision transformer and establishing spatial-temporal decoupling for feature extraction. Although multi-view 3D reconstruction also faces multiple images as input, it cannot immediately inherit their success due to completely ambiguous associations between unstructured views. There is not usable prior relationship, which is similar to the temporally-coherence property in a video. To solve this problem, we propose a novel transformer network for Unstructured Multiple Images (UMIFormer). It exploits transformer blocks for decoupled intra-view encoding and designed blocks for token rectification that mine the correlation between similar tokens from different views to achieve decoupled inter-view encoding. Afterward, all tokens acquired from various branches are compressed into a fixed-size compact representation while preserving rich information for reconstruction by leveraging the similarities between tokens. We empirically demonstrate on ShapeNet and confirm that our decoupled learning method is adaptable for unstructured multiple images. Meanwhile, the experiments also verify our model outperforms existing SOTA methods by a large margin. Code will be available at https://github.com/GaryZhu1996/UMIFormer.

Massive multi-modality datasets play a significant role in facilitating the success of large video-language models. However, current video-language datasets primarily provide text descriptions for visual frames, considering audio to be weakly related information. They usually overlook exploring the potential of inherent audio-visual correlation, leading to monotonous annotation within each modality instead of comprehensive and precise descriptions. Such ignorance results in the difficulty of multiple cross-modality studies. To fulfill this gap, we present MMTrail, a large-scale multi-modality video-language dataset incorporating more than 20M trailer clips with visual captions, and 2M high-quality clips with multimodal captions. Trailers preview full-length video works and integrate context, visual frames, and background music. In particular, the trailer has two main advantages: (1) the topics are diverse, and the content characters are of various types, e.g., film, news, and gaming. (2) the corresponding background music is custom-designed, making it more coherent with the visual context. Upon these insights, we propose a systemic captioning framework, achieving various modality annotations with more than 27.1k hours of trailer videos. Here, to ensure the caption retains music perspective while preserving the authority of visual context, we leverage the advanced LLM to merge all annotations adaptively. In this fashion, our MMtrail dataset potentially paves the path for fine-grained large multimodal-language model training. In experiments, we provide evaluation metrics and benchmark results on our dataset, demonstrating the high quality of our annotation and its effectiveness for model training.

Masked Image Modeling (MIM) has recently been established as a potent pre-training paradigm. A pretext task is constructed by masking patches in an input image, and this masked content is then predicted by a neural network using visible patches as sole input. This pre-training leads to state-of-the-art performance when finetuned for high-level semantic tasks, e.g. image classification and object detection. In this paper we instead seek to learn representations that transfer well to a wide variety of 3D vision and lower-level geometric downstream tasks, such as depth prediction or optical flow estimation. Inspired by MIM, we propose an unsupervised representation learning task trained from pairs of images showing the same scene from different viewpoints. More precisely, we propose the pretext task of cross-view completion where the first input image is partially masked, and this masked content has to be reconstructed from the visible content and the second image. In single-view MIM, the masked content often cannot be inferred precisely from the visible portion only, so the model learns to act as a prior influenced by high-level semantics. In contrast, this ambiguity can be resolved with cross-view completion from the second unmasked image, on the condition that the model is able to understand the spatial relationship between the two images. Our experiments show that our pretext task leads to significantly improved performance for monocular 3D vision downstream tasks such as depth estimation. In addition, our model can be directly applied to binocular downstream tasks like optical flow or relative camera pose estimation, for which we obtain competitive results without bells and whistles, i.e., using a generic architecture without any task-specific design.

Videos are created to express emotion, exchange information, and share experiences. Video synthesis has intrigued researchers for a long time. Despite the rapid progress driven by advances in visual synthesis, most existing studies focus on improving the frames' quality and the transitions between them, while little progress has been made in generating longer videos. In this paper, we present a method that builds on 3D-VQGAN and transformers to generate videos with thousands of frames. Our evaluation shows that our model trained on 16-frame video clips from standard benchmarks such as UCF-101, Sky Time-lapse, and Taichi-HD datasets can generate diverse, coherent, and high-quality long videos. We also showcase conditional extensions of our approach for generating meaningful long videos by incorporating temporal information with text and audio. Videos and code can be found at https://songweige.github.io/projects/tats/index.html.

Reconstructing 4D scenes from video inputs is a crucial yet challenging task. Conventional methods usually rely on the assumptions of multi-view video inputs, known camera parameters, or static scenes, all of which are typically absent under in-the-wild scenarios. In this paper, we relax all these constraints and tackle a highly ambitious but practical task, which we termed as AnyV4D: we assume only one monocular video is available without any camera parameters as input, and we aim to recover the dynamic 4D world alongside the camera poses. To this end, we introduce GFlow, a new framework that utilizes only 2D priors (depth and optical flow) to lift a video (3D) to a 4D explicit representation, entailing a flow of Gaussian splatting through space and time. GFlow first clusters the scene into still and moving parts, then applies a sequential optimization process that optimizes camera poses and the dynamics of 3D Gaussian points based on 2D priors and scene clustering, ensuring fidelity among neighboring points and smooth movement across frames. Since dynamic scenes always introduce new content, we also propose a new pixel-wise densification strategy for Gaussian points to integrate new visual content. Moreover, GFlow transcends the boundaries of mere 4D reconstruction; it also enables tracking of any points across frames without the need for prior training and segments moving objects from the scene in an unsupervised way. Additionally, the camera poses of each frame can be derived from GFlow, allowing for rendering novel views of a video scene through changing camera pose. By employing the explicit representation, we may readily conduct scene-level or object-level editing as desired, underscoring its versatility and power. Visit our project website at: https://littlepure2333.github.io/GFlow

Recent video generative models primarily rely on carefully written text prompts for specific tasks, like inpainting or style editing. They require labor-intensive textual descriptions for input videos, hindering their flexibility to adapt personal/raw videos to user specifications. This paper proposes RACCooN, a versatile and user-friendly video-to-paragraph-to-video generative framework that supports multiple video editing capabilities such as removal, addition, and modification, through a unified pipeline. RACCooN consists of two principal stages: Video-to-Paragraph (V2P) and Paragraph-to-Video (P2V). In the V2P stage, we automatically describe video scenes in well-structured natural language, capturing both the holistic context and focused object details. Subsequently, in the P2V stage, users can optionally refine these descriptions to guide the video diffusion model, enabling various modifications to the input video, such as removing, changing subjects, and/or adding new objects. The proposed approach stands out from other methods through several significant contributions: (1) RACCooN suggests a multi-granular spatiotemporal pooling strategy to generate well-structured video descriptions, capturing both the broad context and object details without requiring complex human annotations, simplifying precise video content editing based on text for users. (2) Our video generative model incorporates auto-generated narratives or instructions to enhance the quality and accuracy of the generated content. It supports the addition of video objects, inpainting, and attribute modification within a unified framework, surpassing existing video editing and inpainting benchmarks. The proposed framework demonstrates impressive versatile capabilities in video-to-paragraph generation, video content editing, and can be incorporated into other SoTA video generative models for further enhancement.

This paper aims to improve the performance of video multimodal large language models (MLLM) via long and rich context (LRC) modeling. As a result, we develop a new version of InternVideo2.5 with a focus on enhancing the original MLLMs' ability to perceive fine-grained details and capture long-form temporal structure in videos. Specifically, our approach incorporates dense vision task annotations into MLLMs using direct preference optimization and develops compact spatiotemporal representations through adaptive hierarchical token compression. Experimental results demonstrate this unique design of LRC greatly improves the results of video MLLM in mainstream video understanding benchmarks (short & long), enabling the MLLM to memorize significantly longer video inputs (at least 6x longer than the original), and master specialized vision capabilities like object tracking and segmentation. Our work highlights the importance of multimodal context richness (length and fineness) in empowering MLLM's innate abilites (focus and memory), providing new insights for future research on video MLLM. Code and models are available at https://github.com/OpenGVLab/InternVideo/tree/main/InternVideo2.5

We consider the task of Image-to-Video (I2V) generation, which involves transforming static images into realistic video sequences based on a textual description. While recent advancements produce photorealistic outputs, they frequently struggle to create videos with accurate and consistent object motion, especially in multi-object scenarios. To address these limitations, we propose a two-stage compositional framework that decomposes I2V generation into: (i) An explicit intermediate representation generation stage, followed by (ii) A video generation stage that is conditioned on this representation. Our key innovation is the introduction of a mask-based motion trajectory as an intermediate representation, that captures both semantic object information and motion, enabling an expressive but compact representation of motion and semantics. To incorporate the learned representation in the second stage, we utilize object-level attention objectives. Specifically, we consider a spatial, per-object, masked-cross attention objective, integrating object-specific prompts into corresponding latent space regions and a masked spatio-temporal self-attention objective, ensuring frame-to-frame consistency for each object. We evaluate our method on challenging benchmarks with multi-object and high-motion scenarios and empirically demonstrate that the proposed method achieves state-of-the-art results in temporal coherence, motion realism, and text-prompt faithfulness. Additionally, we introduce \benchmark, a new challenging benchmark for single-object and multi-object I2V generation, and demonstrate our method's superiority on this benchmark. Project page is available at https://guyyariv.github.io/TTM/.

Though pre-training vision-language models have demonstrated significant benefits in boosting video-text retrieval performance from large-scale web videos, fine-tuning still plays a critical role with manually annotated clips with start and end times, which requires considerable human effort. To address this issue, we explore an alternative cheaper source of annotations, single timestamps, for video-text retrieval. We initialise clips from timestamps in a heuristic way to warm up a retrieval model. Then a video clip editing method is proposed to refine the initial rough boundaries to improve retrieval performance. A student-teacher network is introduced for video clip editing. The teacher model is employed to edit the clips in the training set whereas the student model trains on the edited clips. The teacher weights are updated from the student's after the student's performance increases. Our method is model agnostic and applicable to any retrieval models. We conduct experiments based on three state-of-the-art retrieval models, COOT, VideoCLIP and CLIP4Clip. Experiments conducted on three video retrieval datasets, YouCook2, DiDeMo and ActivityNet-Captions show that our edited clips consistently improve retrieval performance over initial clips across all the three retrieval models.

We explore how reconciling several foundation models (large language models and vision-language models) with a novel unified memory mechanism could tackle the challenging video understanding problem, especially capturing the long-term temporal relations in lengthy videos. In particular, the proposed multimodal agent VideoAgent: 1) constructs a structured memory to store both the generic temporal event descriptions and object-centric tracking states of the video; 2) given an input task query, it employs tools including video segment localization and object memory querying along with other visual foundation models to interactively solve the task, utilizing the zero-shot tool-use ability of LLMs. VideoAgent demonstrates impressive performances on several long-horizon video understanding benchmarks, an average increase of 6.6% on NExT-QA and 26.0% on EgoSchema over baselines, closing the gap between open-sourced models and private counterparts including Gemini 1.5 Pro.

Adapting image models to the video domain has emerged as an efficient paradigm for solving video recognition tasks. Due to the huge number of parameters and effective transferability of image models, performing full fine-tuning is less efficient and even unnecessary. Thus, recent research is shifting its focus toward parameter-efficient image-to-video adaptation. However, these adaptation strategies inevitably introduce extra computational costs to deal with the domain gap and temporal modeling in videos. In this paper, we present a new adaptation paradigm (ZeroI2V) to transfer the image transformers to video recognition tasks (i.e., introduce zero extra cost to the original models during inference). To achieve this goal, we present two core designs. First, to capture the dynamics in videos and reduce the difficulty of image-to-video adaptation, we exploit the flexibility of self-attention and introduce spatial-temporal dual-headed attention (STDHA). This approach efficiently endows the image transformers with temporal modeling capability at zero extra parameters and computation. Second, to handle the domain gap between images and videos, we propose a linear adaption strategy that utilizes lightweight densely placed linear adapters to fully transfer the frozen image models to video recognition. Thanks to the customized linear design, all newly added adapters could be easily merged with the original modules through structural reparameterization after training, enabling zero extra cost during inference. Extensive experiments on representative fully-supervised and few-shot video recognition benchmarks showcase that ZeroI2V can match or even outperform previous state-of-the-art methods while enjoying superior parameter and inference efficiency.

The text-to-video (T2V) generation models, offering convenient visual creation, have recently garnered increasing attention. Despite their substantial potential, the generated videos may present artifacts, including structural implausibility, temporal inconsistency, and a lack of motion, often resulting in near-static video. In this work, we have identified a correlation between the disparity of temporal attention maps across different blocks and the occurrence of temporal inconsistencies. Additionally, we have observed that the energy contained within the temporal attention maps is directly related to the magnitude of motion amplitude in the generated videos. Based on these observations, we present BroadWay, a training-free method to improve the quality of text-to-video generation without introducing additional parameters, augmenting memory or sampling time. Specifically, BroadWay is composed of two principal components: 1) Temporal Self-Guidance improves the structural plausibility and temporal consistency of generated videos by reducing the disparity between the temporal attention maps across various decoder blocks. 2) Fourier-based Motion Enhancement enhances the magnitude and richness of motion by amplifying the energy of the map. Extensive experiments demonstrate that BroadWay significantly improves the quality of text-to-video generation with negligible additional cost.

We present Omni-RGPT, a multimodal large language model designed to facilitate region-level comprehension for both images and videos. To achieve consistent region representation across spatio-temporal dimensions, we introduce Token Mark, a set of tokens highlighting the target regions within the visual feature space. These tokens are directly embedded into spatial regions using region prompts (e.g., boxes or masks) and simultaneously incorporated into the text prompt to specify the target, establishing a direct connection between visual and text tokens. To further support robust video understanding without requiring tracklets, we introduce an auxiliary task that guides Token Mark by leveraging the consistency of the tokens, enabling stable region interpretation across the video. Additionally, we introduce a large-scale region-level video instruction dataset (RegVID-300k). Omni-RGPT achieves state-of-the-art results on image and video-based commonsense reasoning benchmarks while showing strong performance in captioning and referring expression comprehension tasks.

In text-video retrieval, recent works have benefited from the powerful learning capabilities of pre-trained text-image foundation models (e.g., CLIP) by adapting them to the video domain. A critical problem for them is how to effectively capture the rich semantics inside the video using the image encoder of CLIP. To tackle this, state-of-the-art methods adopt complex cross-modal modeling techniques to fuse the text information into video frame representations, which, however, incurs severe efficiency issues in large-scale retrieval systems as the video representations must be recomputed online for every text query. In this paper, we discard this problematic cross-modal fusion process and aim to learn semantically-enhanced representations purely from the video, so that the video representations can be computed offline and reused for different texts. Concretely, we first introduce a spatial-temporal "Prompt Cube" into the CLIP image encoder and iteratively switch it within the encoder layers to efficiently incorporate the global video semantics into frame representations. We then propose to apply an auxiliary video captioning objective to train the frame representations, which facilitates the learning of detailed video semantics by providing fine-grained guidance in the semantic space. With a naive temporal fusion strategy (i.e., mean-pooling) on the enhanced frame representations, we obtain state-of-the-art performances on three benchmark datasets, i.e., MSR-VTT, MSVD, and LSMDC.

Self-supervised representation learning approaches have recently surpassed their supervised learning counterparts on downstream tasks like object detection and image classification. Somewhat mysteriously the recent gains in performance come from training instance classification models, treating each image and it's augmented versions as samples of a single class. In this work, we first present quantitative experiments to demystify these gains. We demonstrate that approaches like MOCO and PIRL learn occlusion-invariant representations. However, they fail to capture viewpoint and category instance invariance which are crucial components for object recognition. Second, we demonstrate that these approaches obtain further gains from access to a clean object-centric training dataset like Imagenet. Finally, we propose an approach to leverage unstructured videos to learn representations that possess higher viewpoint invariance. Our results show that the learned representations outperform MOCOv2 trained on the same data in terms of invariances encoded and the performance on downstream image classification and semantic segmentation tasks.

Video-language pre-training (VLP) has become increasingly important due to its ability to generalize to various vision and language tasks. However, existing egocentric VLP frameworks utilize separate video and language encoders and learn task-specific cross-modal information only during fine-tuning, limiting the development of a unified system. In this work, we introduce the second generation of egocentric video-language pre-training (EgoVLPv2), a significant improvement from the previous generation, by incorporating cross-modal fusion directly into the video and language backbones. EgoVLPv2 learns strong video-text representation during pre-training and reuses the cross-modal attention modules to support different downstream tasks in a flexible and efficient manner, reducing fine-tuning costs. Moreover, our proposed fusion in the backbone strategy is more lightweight and compute-efficient than stacking additional fusion-specific layers. Extensive experiments on a wide range of VL tasks demonstrate the effectiveness of EgoVLPv2 by achieving consistent state-of-the-art performance over strong baselines across all downstream. Our project page can be found at https://shramanpramanick.github.io/EgoVLPv2/.

There has been growing sentiment recently that modern large multimodal models (LMMs) have addressed most of the key challenges related to short video comprehension. As a result, both academia and industry are gradually shifting their attention towards the more complex challenges posed by understanding long-form videos. However, is this really the case? Our studies indicate that LMMs still lack many fundamental reasoning capabilities even when dealing with short videos. We introduce Vinoground, a temporal counterfactual LMM evaluation benchmark encompassing 1000 short and natural video-caption pairs. We demonstrate that existing LMMs severely struggle to distinguish temporal differences between different actions and object transformations. For example, the best model GPT-4o only obtains ~50% on our text and video scores, showing a large gap compared to the human baseline of ~90%. All open-source multimodal models and CLIP-based models perform much worse, producing mostly random chance performance. Through this work, we shed light onto the fact that temporal reasoning in short videos is a problem yet to be fully solved. The dataset and evaluation code are available at https://vinoground.github.io.

Recently, Vision Large Language Models (VLLMs) integrated with vision encoders have shown promising performance in vision understanding. The key of VLLMs is to encode visual content into sequences of visual tokens, enabling VLLMs to simultaneously process both visual and textual content. However, understanding videos, especially long videos, remain a challenge to VLLMs as the number of visual tokens grows rapidly when encoding videos, resulting in the risk of exceeding the context window of VLLMs and introducing heavy computation burden. To restrict the number of visual tokens, existing VLLMs either: (1) uniformly downsample videos into a fixed number of frames or (2) reducing the number of visual tokens encoded from each frame. We argue the former solution neglects the rich temporal cue in videos and the later overlooks the spatial details in each frame. In this work, we present Balanced-VLLM (B-VLLM): a novel VLLM framework that aims to effectively leverage task relevant spatio-temporal cues while restricting the number of visual tokens under the VLLM context window length. At the core of our method, we devise a text-conditioned adaptive frame selection module to identify frames relevant to the visual understanding task. The selected frames are then de-duplicated using a temporal frame token merging technique. The visual tokens of the selected frames are processed through a spatial token sampling module and an optional spatial token merging strategy to achieve precise control over the token count. Experimental results show that B-VLLM is effective in balancing the number of frames and visual tokens in video understanding, yielding superior performance on various video understanding benchmarks. Our code is available at https://github.com/zhuqiangLu/B-VLLM.

Self-supervised monocular depth estimation (SS-MDE) has the potential to scale to vast quantities of data. Unfortunately, existing approaches limit themselves to the automotive domain, resulting in models incapable of generalizing to complex environments such as natural or indoor settings. To address this, we propose a large-scale SlowTV dataset curated from YouTube, containing an order of magnitude more data than existing automotive datasets. SlowTV contains 1.7M images from a rich diversity of environments, such as worldwide seasonal hiking, scenic driving and scuba diving. Using this dataset, we train an SS-MDE model that provides zero-shot generalization to a large collection of indoor/outdoor datasets. The resulting model outperforms all existing SSL approaches and closes the gap on supervised SoTA, despite using a more efficient architecture. We additionally introduce a collection of best-practices to further maximize performance and zero-shot generalization. This includes 1) aspect ratio augmentation, 2) camera intrinsic estimation, 3) support frame randomization and 4) flexible motion estimation. Code is available at https://github.com/jspenmar/slowtv_monodepth.

Recent self-supervised video representation learning methods have found significant success by exploring essential properties of videos, e.g. speed, temporal order, etc. This work exploits an essential yet under-explored property of videos, the video continuity, to obtain supervision signals for self-supervised representation learning. Specifically, we formulate three novel continuity-related pretext tasks, i.e. continuity justification, discontinuity localization, and missing section approximation, that jointly supervise a shared backbone for video representation learning. This self-supervision approach, termed as Continuity Perception Network (CPNet), solves the three tasks altogether and encourages the backbone network to learn local and long-ranged motion and context representations. It outperforms prior arts on multiple downstream tasks, such as action recognition, video retrieval, and action localization. Additionally, the video continuity can be complementary to other coarse-grained video properties for representation learning, and integrating the proposed pretext task to prior arts can yield much performance gains.

We present CLIP2Video network to transfer the image-language pre-training model to video-text retrieval in an end-to-end manner. Leading approaches in the domain of video-and-language learning try to distill the spatio-temporal video features and multi-modal interaction between videos and languages from a large-scale video-text dataset. Different from them, we leverage pretrained image-language model, simplify it as a two-stage framework with co-learning of image-text and enhancing temporal relations between video frames and video-text respectively, make it able to train on comparatively small datasets. Specifically, based on the spatial semantics captured by Contrastive Language-Image Pretraining (CLIP) model, our model involves a Temporal Difference Block to capture motions at fine temporal video frames, and a Temporal Alignment Block to re-align the tokens of video clips and phrases and enhance the multi-modal correlation. We conduct thorough ablation studies, and achieve state-of-the-art performance on major text-to-video and video-to-text retrieval benchmarks, including new records of retrieval accuracy on MSR-VTT, MSVD and VATEX.

Objects in videos are typically characterized by continuous smooth motion. We exploit continuous smooth motion in three ways. 1) Improved accuracy by using object motion as an additional source of supervision, which we obtain by anticipating object locations from a static keyframe. 2) Improved efficiency by only doing the expensive feature computations on a small subset of all frames. Because neighboring video frames are often redundant, we only compute features for a single static keyframe and predict object locations in subsequent frames. 3) Reduced annotation cost, where we only annotate the keyframe and use smooth pseudo-motion between keyframes. We demonstrate computational efficiency, annotation efficiency, and improved mean average precision compared to the state-of-the-art on four datasets: ImageNet VID, EPIC KITCHENS-55, YouTube-BoundingBoxes, and Waymo Open dataset. Our source code is available at https://github.com/L-KID/Videoobject-detection-by-location-anticipation.

The success of monocular depth estimation relies on large and diverse training sets. Due to the challenges associated with acquiring dense ground-truth depth across different environments at scale, a number of datasets with distinct characteristics and biases have emerged. We develop tools that enable mixing multiple datasets during training, even if their annotations are incompatible. In particular, we propose a robust training objective that is invariant to changes in depth range and scale, advocate the use of principled multi-objective learning to combine data from different sources, and highlight the importance of pretraining encoders on auxiliary tasks. Armed with these tools, we experiment with five diverse training datasets, including a new, massive data source: 3D films. To demonstrate the generalization power of our approach we use zero-shot cross-dataset transfer}, i.e. we evaluate on datasets that were not seen during training. The experiments confirm that mixing data from complementary sources greatly improves monocular depth estimation. Our approach clearly outperforms competing methods across diverse datasets, setting a new state of the art for monocular depth estimation. Some results are shown in the supplementary video at https://youtu.be/D46FzVyL9I8

We present a video decomposition method that facilitates layer-based editing of videos with spatiotemporally varying lighting and motion effects. Our neural model decomposes an input video into multiple layered representations, each comprising a 2D texture map, a mask for the original video, and a multiplicative residual characterizing the spatiotemporal variations in lighting conditions. A single edit on the texture maps can be propagated to the corresponding locations in the entire video frames while preserving other contents' consistencies. Our method efficiently learns the layer-based neural representations of a 1080p video in 25s per frame via coordinate hashing and allows real-time rendering of the edited result at 71 fps on a single GPU. Qualitatively, we run our method on various videos to show its effectiveness in generating high-quality editing effects. Quantitatively, we propose to adopt feature-tracking evaluation metrics for objectively assessing the consistency of video editing. Project page: https://lightbulb12294.github.io/hashing-nvd/

Recent video and language pretraining frameworks lack the ability to generate sentences. We present Multimodal Video Generative Pretraining (MV-GPT), a new pretraining framework for learning from unlabelled videos which can be effectively used for generative tasks such as multimodal video captioning. Unlike recent video-language pretraining frameworks, our framework trains both a multimodal video encoder and a sentence decoder jointly. To overcome the lack of captions in unlabelled videos, we leverage the future utterance as an additional text source and propose a bidirectional generation objective -- we generate future utterances given the present mulitmodal context, and also the present utterance given future observations. With this objective, we train an encoder-decoder model end-to-end to generate a caption from raw pixels and transcribed speech directly. Our model achieves state-of-the-art performance for multimodal video captioning on four standard benchmarks, as well as for other video understanding tasks such as VideoQA, video retrieval and action classification.

Methods for image-to-video generation have achieved impressive, photo-realistic quality. However, adjusting specific elements in generated videos, such as object motion or camera movement, is often a tedious process of trial and error, e.g., involving re-generating videos with different random seeds. Recent techniques address this issue by fine-tuning a pre-trained model to follow conditioning signals, such as bounding boxes or point trajectories. Yet, this fine-tuning procedure can be computationally expensive, and it requires datasets with annotated object motion, which can be difficult to procure. In this work, we introduce SG-I2V, a framework for controllable image-to-video generation that is self-guidedx2013offering zero-shot control by relying solely on the knowledge present in a pre-trained image-to-video diffusion model without the need for fine-tuning or external knowledge. Our zero-shot method outperforms unsupervised baselines while being competitive with supervised models in terms of visual quality and motion fidelity.

While Rotary Position Embedding (RoPE) and its variants are widely adopted for their long-context capabilities, the extension of the 1D RoPE to video, with its complex spatio-temporal structure, remains an open challenge. This work first introduces a comprehensive analysis that identifies four key characteristics essential for the effective adaptation of RoPE to video, which have not been fully considered in prior work. As part of our analysis, we introduce a challenging V-NIAH-D (Visual Needle-In-A-Haystack with Distractors) task, which adds periodic distractors into V-NIAH. The V-NIAH-D task demonstrates that previous RoPE variants, lacking appropriate temporal dimension allocation, are easily misled by distractors. Based on our analysis, we introduce VideoRoPE, with a 3D structure designed to preserve spatio-temporal relationships. VideoRoPE features low-frequency temporal allocation to mitigate periodic oscillations, a diagonal layout to maintain spatial symmetry, and adjustable temporal spacing to decouple temporal and spatial indexing. VideoRoPE consistently surpasses previous RoPE variants, across diverse downstream tasks such as long video retrieval, video understanding, and video hallucination. Our code will be available at https://github.com/Wiselnn570/VideoRoPE{https://github.com/Wiselnn570/VideoRoPE}.

In recent years there have been remarkable breakthroughs in image-to-video generation. However, the 3D consistency and camera controllability of generated frames have remained unsolved. Recent studies have attempted to incorporate camera control into the generation process, but their results are often limited to simple trajectories or lack the ability to generate consistent videos from multiple distinct camera paths for the same scene. To address these limitations, we introduce Cavia, a novel framework for camera-controllable, multi-view video generation, capable of converting an input image into multiple spatiotemporally consistent videos. Our framework extends the spatial and temporal attention modules into view-integrated attention modules, improving both viewpoint and temporal consistency. This flexible design allows for joint training with diverse curated data sources, including scene-level static videos, object-level synthetic multi-view dynamic videos, and real-world monocular dynamic videos. To our best knowledge, Cavia is the first of its kind that allows the user to precisely specify camera motion while obtaining object motion. Extensive experiments demonstrate that Cavia surpasses state-of-the-art methods in terms of geometric consistency and perceptual quality. Project Page: https://ir1d.github.io/Cavia/

Text-to-video generative models convert textual prompts into dynamic visual content, offering wide-ranging applications in film production, gaming, and education. However, their real-world performance often falls short of user expectations. One key reason is that these models have not been trained on videos related to some topics users want to create. In this paper, we propose VideoUFO, the first Video dataset specifically curated to align with Users' FOcus in real-world scenarios. Beyond this, our VideoUFO also features: (1) minimal (0.29%) overlap with existing video datasets, and (2) videos searched exclusively via YouTube's official API under the Creative Commons license. These two attributes provide future researchers with greater freedom to broaden their training sources. The VideoUFO comprises over 1.09 million video clips, each paired with both a brief and a detailed caption (description). Specifically, through clustering, we first identify 1,291 user-focused topics from the million-scale real text-to-video prompt dataset, VidProM. Then, we use these topics to retrieve videos from YouTube, split the retrieved videos into clips, and generate both brief and detailed captions for each clip. After verifying the clips with specified topics, we are left with about 1.09 million video clips. Our experiments reveal that (1) current 16 text-to-video models do not achieve consistent performance across all user-focused topics; and (2) a simple model trained on VideoUFO outperforms others on worst-performing topics. The dataset is publicly available at https://huggingface.co/datasets/WenhaoWang/VideoUFO under the CC BY 4.0 License.

The application of Large Vision-Language Models (LVLMs) for analyzing images and videos is an exciting and rapidly evolving field. In recent years, we've seen significant growth in high-quality image-text datasets for fine-tuning image understanding, but there is still a lack of comparable datasets for videos. Additionally, many VideoLLMs are extensions of single-image VLMs, which may not efficiently handle the complexities of longer videos. In this study, we introduce a large-scale synthetic dataset created from proprietary models, using carefully designed prompts to tackle a wide range of questions. We also explore a dynamic visual token compression architecture that strikes a balance between computational efficiency and performance. Our proposed achieves state-of-the-art results across various video tasks and shows impressive generalization, setting new baselines in multi-image understanding. Notably, delivers an absolute improvement of 2.7\% over LLaVA-OneVision on VideoMME and 10.7\% on MuirBench. Codes are available at https://github.com/Hon-Wong/ByteVideoLLM

We present the Decoupled VIdeo Segmentation (DVIS) framework, a novel approach for the challenging task of universal video segmentation, including video instance segmentation (VIS), video semantic segmentation (VSS), and video panoptic segmentation (VPS). Unlike previous methods that model video segmentation in an end-to-end manner, our approach decouples video segmentation into three cascaded sub-tasks: segmentation, tracking, and refinement. This decoupling design allows for simpler and more effective modeling of the spatio-temporal representations of objects, especially in complex scenes and long videos. Accordingly, we introduce two novel components: the referring tracker and the temporal refiner. These components track objects frame by frame and model spatio-temporal representations based on pre-aligned features. To improve the tracking capability of DVIS, we propose a denoising training strategy and introduce contrastive learning, resulting in a more robust framework named DVIS++. Furthermore, we evaluate DVIS++ in various settings, including open vocabulary and using a frozen pre-trained backbone. By integrating CLIP with DVIS++, we present OV-DVIS++, the first open-vocabulary universal video segmentation framework. We conduct extensive experiments on six mainstream benchmarks, including the VIS, VSS, and VPS datasets. Using a unified architecture, DVIS++ significantly outperforms state-of-the-art specialized methods on these benchmarks in both close- and open-vocabulary settings. Code:~https://github.com/zhang-tao-whu/DVIS_Plus.

Current video-based Masked Autoencoders (MAEs) primarily focus on learning effective spatiotemporal representations from a visual perspective, which may lead the model to prioritize general spatial-temporal patterns but often overlook nuanced semantic attributes like specific interactions or sequences that define actions - such as action-specific features that align more closely with human cognition for space-time correspondence. This can limit the model's ability to capture the essence of certain actions that are contextually rich and continuous. Humans are capable of mapping visual concepts, object view invariance, and semantic attributes available in static instances to comprehend natural dynamic scenes or videos. Existing MAEs for videos and static images rely on separate datasets for videos and images, which may lack the rich semantic attributes necessary for fully understanding the learned concepts, especially when compared to using video and corresponding sampled frame images together. To this end, we propose CrossVideoMAE an end-to-end self-supervised cross-modal contrastive learning MAE that effectively learns both video-level and frame-level rich spatiotemporal representations and semantic attributes. Our method integrates mutual spatiotemporal information from videos with spatial information from sampled frames within a feature-invariant space, while encouraging invariance to augmentations within the video domain. This objective is achieved through jointly embedding features of visible tokens and combining feature correspondence within and across modalities, which is critical for acquiring rich, label-free guiding signals from both video and frame image modalities in a self-supervised manner. Extensive experiments demonstrate that our approach surpasses previous state-of-the-art methods and ablation studies validate the effectiveness of our approach.

Multi-task approaches to joint depth and segmentation prediction are well-studied for monocular images. Yet, predictions from a single-view are inherently limited, while multiple views are available in many robotics applications. On the other end of the spectrum, video-based and full 3D methods require numerous frames to perform reconstruction and segmentation. With this work we propose a Multi-View Stereo (MVS) technique for depth prediction that benefits from rich semantic features of the Segment Anything Model (SAM). This enhanced depth prediction, in turn, serves as a prompt to our Transformer-based semantic segmentation decoder. We report the mutual benefit that both tasks enjoy in our quantitative and qualitative studies on the ScanNet dataset. Our approach consistently outperforms single-task MVS and segmentation models, along with multi-task monocular methods.

Advancements in Large Language Models (LLMs) inspire various strategies for integrating video modalities. A key approach is Video-LLMs, which incorporate an optimizable interface linking sophisticated video encoders to LLMs. However, due to computation and data limitations, these Video-LLMs are typically pre-trained to process only short videos, limiting their broader application for understanding longer video content. Additionally, fine-tuning Video-LLMs to handle longer videos is cost-prohibitive. Consequently, it becomes essential to explore the interpolation of Video-LLMs under a completely training-free setting. In this paper, we first identify the primary challenges in interpolating Video-LLMs: (1) the video encoder and modality alignment projector are fixed, preventing the integration of additional frames into Video-LLMs, and (2) the LLM backbone is limited in its content length capabilities, which complicates the processing of an increased number of video tokens. To address these challenges, we propose a specific INTerPolation method for Video-LLMs (INTP-Video-LLMs). We introduce an alternative video token rearrangement technique that circumvents limitations imposed by the fixed video encoder and alignment projector. Furthermore, we introduce a training-free LLM context window extension method to enable Video-LLMs to understand a correspondingly increased number of visual tokens.

We present a foundation model for zero-shot metric monocular depth estimation. Our model, Depth Pro, synthesizes high-resolution depth maps with unparalleled sharpness and high-frequency details. The predictions are metric, with absolute scale, without relying on the availability of metadata such as camera intrinsics. And the model is fast, producing a 2.25-megapixel depth map in 0.3 seconds on a standard GPU. These characteristics are enabled by a number of technical contributions, including an efficient multi-scale vision transformer for dense prediction, a training protocol that combines real and synthetic datasets to achieve high metric accuracy alongside fine boundary tracing, dedicated evaluation metrics for boundary accuracy in estimated depth maps, and state-of-the-art focal length estimation from a single image. Extensive experiments analyze specific design choices and demonstrate that Depth Pro outperforms prior work along multiple dimensions. We release code and weights at https://github.com/apple/ml-depth-pro

In a video, an object usually appears as the trajectory, i.e., it spans over a few spatial but longer temporal patches, that contains abundant spatiotemporal contexts. However, modern Video-Language BERTs (VDL-BERTs) neglect this trajectory characteristic that they usually follow image-language BERTs (IL-BERTs) to deploy the patch-to-word (P2W) attention that may over-exploit trivial spatial contexts and neglect significant temporal contexts. To amend this, we propose a novel TW-BERT to learn Trajectory-Word alignment by a newly designed trajectory-to-word (T2W) attention for solving video-language tasks. Moreover, previous VDL-BERTs usually uniformly sample a few frames into the model while different trajectories have diverse graininess, i.e., some trajectories span longer frames and some span shorter, and using a few frames will lose certain useful temporal contexts. However, simply sampling more frames will also make pre-training infeasible due to the largely increased training burdens. To alleviate the problem, during the fine-tuning stage, we insert a novel Hierarchical Frame-Selector (HFS) module into the video encoder. HFS gradually selects the suitable frames conditioned on the text context for the later cross-modal encoder to learn better trajectory-word alignments. By the proposed T2W attention and HFS, our TW-BERT achieves SOTA performances on text-to-video retrieval tasks, and comparable performances on video question-answering tasks with some VDL-BERTs trained on much more data. The code will be available in the supplementary material.

Given a video and a set of input object masks, an omnimatte method aims to decompose the video into semantically meaningful layers containing individual objects along with their associated effects, such as shadows and reflections. Existing omnimatte methods assume a static background or accurate pose and depth estimation and produce poor decompositions when these assumptions are violated. Furthermore, due to the lack of generative prior on natural videos, existing methods cannot complete dynamic occluded regions. We present a novel generative layered video decomposition framework to address the omnimatte problem. Our method does not assume a stationary scene or require camera pose or depth information and produces clean, complete layers, including convincing completions of occluded dynamic regions. Our core idea is to train a video diffusion model to identify and remove scene effects caused by a specific object. We show that this model can be finetuned from an existing video inpainting model with a small, carefully curated dataset, and demonstrate high-quality decompositions and editing results for a wide range of casually captured videos containing soft shadows, glossy reflections, splashing water, and more.

Learning visual knowledge from massive weakly-labeled web videos has attracted growing research interests thanks to the large corpus of easily accessible video data on the Internet. However, for video action recognition, the action of interest might only exist in arbitrary clips of untrimmed web videos, resulting in high label noises in the temporal space. To address this issue, we introduce a new method for pre-training video action recognition models using queried web videos. Instead of trying to filter out, we propose to convert the potential noises in these queried videos to useful supervision signals by defining the concept of Sub-Pseudo Label (SPL). Specifically, SPL spans out a new set of meaningful "middle ground" label space constructed by extrapolating the original weak labels during video querying and the prior knowledge distilled from a teacher model. Consequently, SPL provides enriched supervision for video models to learn better representations. SPL is fairly simple and orthogonal to popular teacher-student self-training frameworks without extra training cost. We validate the effectiveness of our method on four video action recognition datasets and a weakly-labeled image dataset to study the generalization ability. Experiments show that SPL outperforms several existing pre-training strategies using pseudo-labels and the learned representations lead to competitive results when fine-tuning on HMDB-51 and UCF-101 compared with recent pre-training methods.

In recent years, considerable progress on the task of text-video retrieval has been achieved by leveraging large-scale pretraining on visual and audio datasets to construct powerful video encoders. By contrast, despite the natural symmetry, the design of effective algorithms for exploiting large-scale language pretraining remains under-explored. In this work, we are the first to investigate the design of such algorithms and propose a novel generalized distillation method, TeachText, which leverages complementary cues from multiple text encoders to provide an enhanced supervisory signal to the retrieval model. Moreover, we extend our method to video side modalities and show that we can effectively reduce the number of used modalities at test time without compromising performance. Our approach advances the state of the art on several video retrieval benchmarks by a significant margin and adds no computational overhead at test time. Last but not least, we show an effective application of our method for eliminating noise from retrieval datasets. Code and data can be found at https://www.robots.ox.ac.uk/~vgg/research/teachtext/.

The quality of the data and annotation upper-bounds the quality of a downstream model. While there exist large text corpora and image-text pairs, high-quality video-text data is much harder to collect. First of all, manual labeling is more time-consuming, as it requires an annotator to watch an entire video. Second, videos have a temporal dimension, consisting of several scenes stacked together, and showing multiple actions. Accordingly, to establish a video dataset with high-quality captions, we propose an automatic approach leveraging multimodal inputs, such as textual video description, subtitles, and individual video frames. Specifically, we curate 3.8M high-resolution videos from the publicly available HD-VILA-100M dataset. We then split them into semantically consistent video clips, and apply multiple cross-modality teacher models to obtain captions for each video. Next, we finetune a retrieval model on a small subset where the best caption of each video is manually selected and then employ the model in the whole dataset to select the best caption as the annotation. In this way, we get 70M videos paired with high-quality text captions. We dub the dataset as Panda-70M. We show the value of the proposed dataset on three downstream tasks: video captioning, video and text retrieval, and text-driven video generation. The models trained on the proposed data score substantially better on the majority of metrics across all the tasks.

We extend multimodal transformers to include 3D camera motion as a conditioning signal for the task of video generation. Generative video models are becoming increasingly powerful, thus focusing research efforts on methods of controlling the output of such models. We propose to add virtual 3D camera controls to generative video methods by conditioning generated video on an encoding of three-dimensional camera movement over the course of the generated video. Results demonstrate that we are (1) able to successfully control the camera during video generation, starting from a single frame and a camera signal, and (2) we demonstrate the accuracy of the generated 3D camera paths using traditional computer vision methods.

We propose VidStyleODE, a spatiotemporally continuous disentangled Video representation based upon StyleGAN and Neural-ODEs. Effective traversal of the latent space learned by Generative Adversarial Networks (GANs) has been the basis for recent breakthroughs in image editing. However, the applicability of such advancements to the video domain has been hindered by the difficulty of representing and controlling videos in the latent space of GANs. In particular, videos are composed of content (i.e., appearance) and complex motion components that require a special mechanism to disentangle and control. To achieve this, VidStyleODE encodes the video content in a pre-trained StyleGAN W_+ space and benefits from a latent ODE component to summarize the spatiotemporal dynamics of the input video. Our novel continuous video generation process then combines the two to generate high-quality and temporally consistent videos with varying frame rates. We show that our proposed method enables a variety of applications on real videos: text-guided appearance manipulation, motion manipulation, image animation, and video interpolation and extrapolation. Project website: https://cyberiada.github.io/VidStyleODE

Recent years have witnessed a big convergence of language, vision, and multi-modal pretraining. In this work, we present mPLUG-2, a new unified paradigm with modularized design for multi-modal pretraining, which can benefit from modality collaboration while addressing the problem of modality entanglement. In contrast to predominant paradigms of solely relying on sequence-to-sequence generation or encoder-based instance discrimination, mPLUG-2 introduces a multi-module composition network by sharing common universal modules for modality collaboration and disentangling different modality modules to deal with modality entanglement. It is flexible to select different modules for different understanding and generation tasks across all modalities including text, image, and video. Empirical study shows that mPLUG-2 achieves state-of-the-art or competitive results on a broad range of over 30 downstream tasks, spanning multi-modal tasks of image-text and video-text understanding and generation, and uni-modal tasks of text-only, image-only, and video-only understanding. Notably, mPLUG-2 shows new state-of-the-art results of 48.0 top-1 accuracy and 80.3 CIDEr on the challenging MSRVTT video QA and video caption tasks with a far smaller model size and data scale. It also demonstrates strong zero-shot transferability on vision-language and video-language tasks. Code and models will be released in https://github.com/alibaba/AliceMind.

We propose MAMo, a novel memory and attention frame-work for monocular video depth estimation. MAMo can augment and improve any single-image depth estimation networks into video depth estimation models, enabling them to take advantage of the temporal information to predict more accurate depth. In MAMo, we augment model with memory which aids the depth prediction as the model streams through the video. Specifically, the memory stores learned visual and displacement tokens of the previous time instances. This allows the depth network to cross-reference relevant features from the past when predicting depth on the current frame. We introduce a novel scheme to continuously update the memory, optimizing it to keep tokens that correspond with both the past and the present visual information. We adopt attention-based approach to process memory features where we first learn the spatio-temporal relation among the resultant visual and displacement memory tokens using self-attention module. Further, the output features of self-attention are aggregated with the current visual features through cross-attention. The cross-attended features are finally given to a decoder to predict depth on the current frame. Through extensive experiments on several benchmarks, including KITTI, NYU-Depth V2, and DDAD, we show that MAMo consistently improves monocular depth estimation networks and sets new state-of-the-art (SOTA) accuracy. Notably, our MAMo video depth estimation provides higher accuracy with lower latency, when omparing to SOTA cost-volume-based video depth models.

Masked autoencoding has shown excellent performance on self-supervised video representation learning. Temporal redundancy has led to a high masking ratio and customized masking strategy in VideoMAE. In this paper, we aim to further improve the performance of video masked autoencoding by introducing a motion guided masking strategy. Our key insight is that motion is a general and unique prior in video, which should be taken into account during masked pre-training. Our motion guided masking explicitly incorporates motion information to build temporal consistent masking volume. Based on this masking volume, we can track the unmasked tokens in time and sample a set of temporal consistent cubes from videos. These temporal aligned unmasked tokens will further relieve the information leakage issue in time and encourage the MGMAE to learn more useful structure information. We implement our MGMAE with an online efficient optical flow estimator and backward masking map warping strategy. We perform experiments on the datasets of Something-Something V2 and Kinetics-400, demonstrating the superior performance of our MGMAE to the original VideoMAE. In addition, we provide the visualization analysis to illustrate that our MGMAE can sample temporal consistent cubes in a motion-adaptive manner for more effective video pre-training.

Large-scale vision-language models (VLM) have shown impressive results for language-guided search applications. While these models allow category-level queries, they currently struggle with personalized searches for moments in a video where a specific object instance such as ``My dog Biscuit'' appears. We present the following three contributions to address this problem. First, we describe a method to meta-personalize a pre-trained VLM, i.e., learning how to learn to personalize a VLM at test time to search in video. Our method extends the VLM's token vocabulary by learning novel word embeddings specific to each instance. To capture only instance-specific features, we represent each instance embedding as a combination of shared and learned global category features. Second, we propose to learn such personalization without explicit human supervision. Our approach automatically identifies moments of named visual instances in video using transcripts and vision-language similarity in the VLM's embedding space. Finally, we introduce This-Is-My, a personal video instance retrieval benchmark. We evaluate our approach on This-Is-My and DeepFashion2 and show that we obtain a 15% relative improvement over the state of the art on the latter dataset.

Recent advancements in large-scale video-language models have shown significant potential for real-time planning and detailed interactions. However, their high computational demands and the scarcity of annotated datasets limit their practicality for academic researchers. In this work, we introduce VideoLLaMB, a novel framework that utilizes temporal memory tokens within bridge layers to allow for the encoding of entire video sequences alongside historical visual data, effectively preserving semantic continuity and enhancing model performance across various tasks. This approach includes recurrent memory tokens and a SceneTilling algorithm, which segments videos into independent semantic units to preserve semantic integrity. Empirically, VideoLLaMB significantly outstrips existing video-language models, demonstrating a 5.5 points improvement over its competitors across three VideoQA benchmarks, and 2.06 points on egocentric planning. Comprehensive results on the MVBench show that VideoLLaMB-7B achieves markedly better results than previous 7B models of same LLM. Remarkably, it maintains robust performance as PLLaVA even as video length increases up to 8 times. Besides, the frame retrieval results on our specialized Needle in a Video Haystack (NIAVH) benchmark, further validate VideoLLaMB's prowess in accurately identifying specific frames within lengthy videos. Our SceneTilling algorithm also enables the generation of streaming video captions directly, without necessitating additional training. In terms of efficiency, VideoLLaMB, trained on 16 frames, supports up to 320 frames on a single Nvidia A100 GPU with linear GPU memory scaling, ensuring both high performance and cost-effectiveness, thereby setting a new foundation for long-form video-language models in both academic and practical applications.

As Artificial Intelligence Generated Content (AIGC) advances, a variety of methods have been developed to generate text, images, videos, and 3D objects from single or multimodal inputs, contributing efforts to emulate human-like cognitive content creation. However, generating realistic large-scale scenes from a single input presents a challenge due to the complexities involved in ensuring consistency across extrapolated views generated by models. Benefiting from recent video generation models and implicit neural representations, we propose Scene123, a 3D scene generation model, that not only ensures realism and diversity through the video generation framework but also uses implicit neural fields combined with Masked Autoencoders (MAE) to effectively ensures the consistency of unseen areas across views. Specifically, we initially warp the input image (or an image generated from text) to simulate adjacent views, filling the invisible areas with the MAE model. However, these filled images usually fail to maintain view consistency, thus we utilize the produced views to optimize a neural radiance field, enhancing geometric consistency. Moreover, to further enhance the details and texture fidelity of generated views, we employ a GAN-based Loss against images derived from the input image through the video generation model. Extensive experiments demonstrate that our method can generate realistic and consistent scenes from a single prompt. Both qualitative and quantitative results indicate that our approach surpasses existing state-of-the-art methods. We show encourage video examples at https://yiyingyang12.github.io/Scene123.github.io/.

Dynamic novel view synthesis aims to capture the temporal evolution of visual content within videos. Existing methods struggle to distinguishing between motion and structure, particularly in scenarios where camera poses are either unknown or constrained compared to object motion. Furthermore, with information solely from reference images, it is extremely challenging to hallucinate unseen regions that are occluded or partially observed in the given videos. To address these issues, we first finetune a pretrained RGB-D diffusion model on the video frames using a customization technique. Subsequently, we distill the knowledge from the finetuned model to a 4D representations encompassing both dynamic and static Neural Radiance Fields (NeRF) components. The proposed pipeline achieves geometric consistency while preserving the scene identity. We perform thorough experiments to evaluate the efficacy of the proposed method qualitatively and quantitatively. Our results demonstrate the robustness and utility of our approach in challenging cases, further advancing dynamic novel view synthesis.

Recent advancements in diffusion-based video generation have showcased remarkable results, yet the gap between synthetic and real-world videos remains under-explored. In this study, we examine this gap from three fundamental perspectives: appearance, motion, and geometry, comparing real-world videos with those generated by a state-of-the-art AI model, Stable Video Diffusion. To achieve this, we train three classifiers using 3D convolutional networks, each targeting distinct aspects: vision foundation model features for appearance, optical flow for motion, and monocular depth for geometry. Each classifier exhibits strong performance in fake video detection, both qualitatively and quantitatively. This indicates that AI-generated videos are still easily detectable, and a significant gap between real and fake videos persists. Furthermore, utilizing the Grad-CAM, we pinpoint systematic failures of AI-generated videos in appearance, motion, and geometry. Finally, we propose an Ensemble-of-Experts model that integrates appearance, optical flow, and depth information for fake video detection, resulting in enhanced robustness and generalization ability. Our model is capable of detecting videos generated by Sora with high accuracy, even without exposure to any Sora videos during training. This suggests that the gap between real and fake videos can be generalized across various video generative models. Project page: https://justin-crchang.github.io/3DCNNDetection.github.io/

Generating fine-grained video descriptions is a fundamental challenge in video understanding. In this work, we introduce Tarsier, a family of large-scale video-language models designed to generate high-quality video descriptions. Tarsier employs CLIP-ViT to encode frames separately and then uses an LLM to model temporal relationships. Despite its simple architecture, we demonstrate that with a meticulously designed two-stage training procedure, the Tarsier models exhibit substantially stronger video description capabilities than any existing open-source model, showing a +51.4% advantage in human side-by-side evaluation over the strongest model. Additionally, they are comparable to state-of-the-art proprietary models, with a +12.3% advantage against GPT-4V and a -6.7% disadvantage against Gemini 1.5 Pro. Besides video description, Tarsier proves to be a versatile generalist model, achieving new state-of-the-art results across nine public benchmarks, including multi-choice VQA, open-ended VQA, and zero-shot video captioning. Our second contribution is the introduction of a new benchmark for evaluating video description models, consisting of a new challenging dataset featuring videos from diverse sources and varying complexity, along with an automatic method specifically designed to assess the quality of fine-grained video descriptions. We make our models and evaluation benchmark publicly available at https://github.com/bytedance/tarsier.

Recent advancements in static feed-forward scene reconstruction have demonstrated significant progress in high-quality novel view synthesis. However, these models often struggle with generalizability across diverse environments and fail to effectively handle dynamic content. We present BTimer (short for BulletTimer), the first motion-aware feed-forward model for real-time reconstruction and novel view synthesis of dynamic scenes. Our approach reconstructs the full scene in a 3D Gaussian Splatting representation at a given target ('bullet') timestamp by aggregating information from all the context frames. Such a formulation allows BTimer to gain scalability and generalization by leveraging both static and dynamic scene datasets. Given a casual monocular dynamic video, BTimer reconstructs a bullet-time scene within 150ms while reaching state-of-the-art performance on both static and dynamic scene datasets, even compared with optimization-based approaches.

We introduce a novel approach that takes a single semantic mask as input to synthesize multi-view consistent color images of natural scenes, trained with a collection of single images from the Internet. Prior works on 3D-aware image synthesis either require multi-view supervision or learning category-level prior for specific classes of objects, which can hardly work for natural scenes. Our key idea to solve this challenging problem is to use a semantic field as the intermediate representation, which is easier to reconstruct from an input semantic mask and then translate to a radiance field with the assistance of off-the-shelf semantic image synthesis models. Experiments show that our method outperforms baseline methods and produces photorealistic, multi-view consistent videos of a variety of natural scenes.

This paper presents a novel framework for converting 2D videos to immersive stereoscopic 3D, addressing the growing demand for 3D content in immersive experience. Leveraging foundation models as priors, our approach overcomes the limitations of traditional methods and boosts the performance to ensure the high-fidelity generation required by the display devices. The proposed system consists of two main steps: depth-based video splatting for warping and extracting occlusion mask, and stereo video inpainting. We utilize pre-trained stable video diffusion as the backbone and introduce a fine-tuning protocol for the stereo video inpainting task. To handle input video with varying lengths and resolutions, we explore auto-regressive strategies and tiled processing. Finally, a sophisticated data processing pipeline has been developed to reconstruct a large-scale and high-quality dataset to support our training. Our framework demonstrates significant improvements in 2D-to-3D video conversion, offering a practical solution for creating immersive content for 3D devices like Apple Vision Pro and 3D displays. In summary, this work contributes to the field by presenting an effective method for generating high-quality stereoscopic videos from monocular input, potentially transforming how we experience digital media.

In this paper, we propose a framework named OCSampler to explore a compact yet effective video representation with one short clip for efficient video recognition. Recent works prefer to formulate frame sampling as a sequential decision task by selecting frames one by one according to their importance, while we present a new paradigm of learning instance-specific video condensation policies to select informative frames for representing the entire video only in a single step. Our basic motivation is that the efficient video recognition task lies in processing a whole sequence at once rather than picking up frames sequentially. Accordingly, these policies are derived from a light-weighted skim network together with a simple yet effective policy network within one step. Moreover, we extend the proposed method with a frame number budget, enabling the framework to produce correct predictions in high confidence with as few frames as possible. Experiments on four benchmarks, i.e., ActivityNet, Mini-Kinetics, FCVID, Mini-Sports1M, demonstrate the effectiveness of our OCSampler over previous methods in terms of accuracy, theoretical computational expense, actual inference speed. We also evaluate its generalization power across different classifiers, sampled frames, and search spaces. Especially, we achieve 76.9% mAP and 21.7 GFLOPs on ActivityNet with an impressive throughput: 123.9 Videos/s on a single TITAN Xp GPU.

Despite the rapid integration of video perception capabilities into Large Multimodal Models (LMMs), the underlying mechanisms driving their video understanding remain poorly understood. Consequently, many design decisions in this domain are made without proper justification or analysis. The high computational cost of training and evaluating such models, coupled with limited open research, hinders the development of video-LMMs. To address this, we present a comprehensive study that helps uncover what effectively drives video understanding in LMMs. We begin by critically examining the primary contributors to the high computational requirements associated with video-LMM research and discover Scaling Consistency, wherein design and training decisions made on smaller models and datasets (up to a critical size) effectively transfer to larger models. Leveraging these insights, we explored many video-specific aspects of video-LMMs, including video sampling, architectures, data composition, training schedules, and more. For example, we demonstrated that fps sampling during training is vastly preferable to uniform frame sampling and which vision encoders are the best for video representation. Guided by these findings, we introduce Apollo, a state-of-the-art family of LMMs that achieve superior performance across different model sizes. Our models can perceive hour-long videos efficiently, with Apollo-3B outperforming most existing 7B models with an impressive 55.1 on LongVideoBench. Apollo-7B is state-of-the-art compared to 7B LMMs with a 70.9 on MLVU, and 63.3 on Video-MME.

Recent advances in video-based multimodal large language models (Video-LLMs) have significantly improved video understanding by processing videos as sequences of image frames. However, many existing methods treat frames independently in the vision backbone, lacking explicit temporal modeling, which limits their ability to capture dynamic patterns and efficiently handle long videos. To address these limitations, we introduce STORM (Spatiotemporal TOken Reduction for Multimodal LLMs), a novel architecture incorporating a dedicated temporal encoder between the image encoder and the LLM. Our temporal encoder leverages the Mamba State Space Model to integrate temporal information into image tokens, generating enriched representations that preserve inter-frame dynamics across the entire video sequence. This enriched encoding not only enhances video reasoning capabilities but also enables effective token reduction strategies, including test-time sampling and training-based temporal and spatial pooling, substantially reducing computational demands on the LLM without sacrificing key temporal information. By integrating these techniques, our approach simultaneously reduces training and inference latency while improving performance, enabling efficient and robust video understanding over extended temporal contexts. Extensive evaluations show that STORM achieves state-of-the-art results across various long video understanding benchmarks (more than 5\% improvement on MLVU and LongVideoBench) while reducing the computation costs by up to 8times and the decoding latency by 2.4-2.9times for the fixed numbers of input frames. Project page is available at https://research.nvidia.com/labs/lpr/storm

This paper presents WALDO (WArping Layer-Decomposed Objects), a novel approach to the prediction of future video frames from past ones. Individual images are decomposed into multiple layers combining object masks and a small set of control points. The layer structure is shared across all frames in each video to build dense inter-frame connections. Complex scene motions are modeled by combining parametric geometric transformations associated with individual layers, and video synthesis is broken down into discovering the layers associated with past frames, predicting the corresponding transformations for upcoming ones and warping the associated object regions accordingly, and filling in the remaining image parts. Extensive experiments on multiple benchmarks including urban videos (Cityscapes and KITTI) and videos featuring nonrigid motions (UCF-Sports and H3.6M), show that our method consistently outperforms the state of the art by a significant margin in every case. Code, pretrained models, and video samples synthesized by our approach can be found in the project webpage https://16lemoing.github.io/waldo.

Creating a vivid video from the event or scenario in our imagination is a truly fascinating experience. Recent advancements in text-to-video synthesis have unveiled the potential to achieve this with prompts only. While text is convenient in conveying the overall scene context, it may be insufficient to control precisely. In this paper, we explore customized video generation by utilizing text as context description and motion structure (e.g. frame-wise depth) as concrete guidance. Our method, dubbed Make-Your-Video, involves joint-conditional video generation using a Latent Diffusion Model that is pre-trained for still image synthesis and then promoted for video generation with the introduction of temporal modules. This two-stage learning scheme not only reduces the computing resources required, but also improves the performance by transferring the rich concepts available in image datasets solely into video generation. Moreover, we use a simple yet effective causal attention mask strategy to enable longer video synthesis, which mitigates the potential quality degradation effectively. Experimental results show the superiority of our method over existing baselines, particularly in terms of temporal coherence and fidelity to users' guidance. In addition, our model enables several intriguing applications that demonstrate potential for practical usage.

We present an approach to estimating camera rotation in crowded, real-world scenes from handheld monocular video. While camera rotation estimation is a well-studied problem, no previous methods exhibit both high accuracy and acceptable speed in this setting. Because the setting is not addressed well by other datasets, we provide a new dataset and benchmark, with high-accuracy, rigorously verified ground truth, on 17 video sequences. Methods developed for wide baseline stereo (e.g., 5-point methods) perform poorly on monocular video. On the other hand, methods used in autonomous driving (e.g., SLAM) leverage specific sensor setups, specific motion models, or local optimization strategies (lagging batch processing) and do not generalize well to handheld video. Finally, for dynamic scenes, commonly used robustification techniques like RANSAC require large numbers of iterations, and become prohibitively slow. We introduce a novel generalization of the Hough transform on SO(3) to efficiently and robustly find the camera rotation most compatible with optical flow. Among comparably fast methods, ours reduces error by almost 50\% over the next best, and is more accurate than any method, irrespective of speed. This represents a strong new performance point for crowded scenes, an important setting for computer vision. The code and the dataset are available at https://fabiendelattre.com/robust-rotation-estimation.

Recent text-to-video diffusion models have achieved impressive progress. In practice, users often desire the ability to control object motion and camera movement independently for customized video creation. However, current methods lack the focus on separately controlling object motion and camera movement in a decoupled manner, which limits the controllability and flexibility of text-to-video models. In this paper, we introduce Direct-a-Video, a system that allows users to independently specify motions for one or multiple objects and/or camera movements, as if directing a video. We propose a simple yet effective strategy for the decoupled control of object motion and camera movement. Object motion is controlled through spatial cross-attention modulation using the model's inherent priors, requiring no additional optimization. For camera movement, we introduce new temporal cross-attention layers to interpret quantitative camera movement parameters. We further employ an augmentation-based approach to train these layers in a self-supervised manner on a small-scale dataset, eliminating the need for explicit motion annotation. Both components operate independently, allowing individual or combined control, and can generalize to open-domain scenarios. Extensive experiments demonstrate the superiority and effectiveness of our method. Project page: https://direct-a-video.github.io/.

Large-scale image-language pretrained models, e.g., CLIP, have demonstrated remarkable proficiency in acquiring general multi-modal knowledge through web-scale image-text data. Despite the impressive performance of image-language models on various image tasks, how to effectively expand them on general video understanding remains an area of ongoing exploration. In this paper, we investigate the image-to-video transferring from the perspective of the model and the data, unveiling two key obstacles impeding the adaptation of image-language models: non-generalizable temporal modeling and partially misaligned video-text data. To address these challenges, we propose Spatial-Temporal Auxiliary Network with Mutual-guided alignment module (Mug-STAN), a simple yet effective framework extending image-text model to diverse video tasks and video-text data.Specifically, STAN adopts a branch structure with decomposed spatial-temporal modules to enable generalizable temporal modeling, while Mug suppresses misalignment by introducing token-wise feature aggregation of either modality from the other. Extensive experimental results verify Mug-STAN significantly improves adaptation of language-image pretrained models such as CLIP and CoCa at both video-text post-pretraining and finetuning stages. With our solution, state-of-the-art zero-shot and finetuning results on various downstream datasets, including MSR-VTT, DiDeMo, LSMDC, Kinetics-400, Something-Something-2, HMDB-51, UCF- 101, and AVA, are achieved. Moreover, by integrating pretrained Mug-STAN with the emerging multimodal dialogue model, we can realize zero-shot video chatting. Codes are available at https://github.com/farewellthree/STAN

Reconstructing accurate 3D scenes from images is a long-standing vision task. Due to the ill-posedness of the single-image reconstruction problem, most well-established methods are built upon multi-view geometry. State-of-the-art (SOTA) monocular metric depth estimation methods can only handle a single camera model and are unable to perform mixed-data training due to the metric ambiguity. Meanwhile, SOTA monocular methods trained on large mixed datasets achieve zero-shot generalization by learning affine-invariant depths, which cannot recover real-world metrics. In this work, we show that the key to a zero-shot single-view metric depth model lies in the combination of large-scale data training and resolving the metric ambiguity from various camera models. We propose a canonical camera space transformation module, which explicitly addresses the ambiguity problems and can be effortlessly plugged into existing monocular models. Equipped with our module, monocular models can be stably trained with over 8 million images with thousands of camera models, resulting in zero-shot generalization to in-the-wild images with unseen camera settings. Experiments demonstrate SOTA performance of our method on 7 zero-shot benchmarks. Notably, our method won the championship in the 2nd Monocular Depth Estimation Challenge. Our method enables the accurate recovery of metric 3D structures on randomly collected internet images, paving the way for plausible single-image metrology. The potential benefits extend to downstream tasks, which can be significantly improved by simply plugging in our model. For example, our model relieves the scale drift issues of monocular-SLAM (Fig. 1), leading to high-quality metric scale dense mapping. The code is available at https://github.com/YvanYin/Metric3D.

The success of Multimodal Large Language Models (MLLMs) in the image domain has garnered wide attention from the research community. Drawing on previous successful experiences, researchers have recently explored extending the success to the video understanding realms. Apart from training from scratch, an efficient way is to utilize the pre-trained image-LLMs, leading to two mainstream approaches, i.e. zero-shot inference and further fine-tuning with video data. In this work, our study of these approaches harvests an effective data augmentation method. We first make a deeper inspection of the zero-shot inference way and identify two limitations, i.e. limited generalization and lack of temporal understanding capabilities. Thus, we further investigate the fine-tuning approach and find a low learning efficiency when simply using all the video data samples, which can be attributed to a lack of instruction diversity. Aiming at this issue, we develop a method called T2Vid to synthesize video-like samples to enrich the instruction diversity in the training corpus. Integrating these data enables a simple and efficient training scheme, which achieves performance comparable to or even superior to using full video datasets by training with just 15% the sample size. Meanwhile, we find that the proposed scheme can boost the performance of long video understanding without training with long video samples. We hope our study will spark more thinking about using MLLMs for video understanding and curation of high-quality data. The code is released at https://github.com/xjtupanda/T2Vid.

We tackle the task of reconstructing hand-object interactions from short video clips. Given an input video, our approach casts 3D inference as a per-video optimization and recovers a neural 3D representation of the object shape, as well as the time-varying motion and hand articulation. While the input video naturally provides some multi-view cues to guide 3D inference, these are insufficient on their own due to occlusions and limited viewpoint variations. To obtain accurate 3D, we augment the multi-view signals with generic data-driven priors to guide reconstruction. Specifically, we learn a diffusion network to model the conditional distribution of (geometric) renderings of objects conditioned on hand configuration and category label, and leverage it as a prior to guide the novel-view renderings of the reconstructed scene. We empirically evaluate our approach on egocentric videos across 6 object categories, and observe significant improvements over prior single-view and multi-view methods. Finally, we demonstrate our system's ability to reconstruct arbitrary clips from YouTube, showing both 1st and 3rd person interactions.

Video descriptions are crucial for blind and low vision (BLV) users to access visual content. However, current artificial intelligence models for generating descriptions often fall short due to limitations in the quality of human annotations within training datasets, resulting in descriptions that do not fully meet BLV users' needs. To address this gap, we introduce VideoA11y, an approach that leverages multimodal large language models (MLLMs) and video accessibility guidelines to generate descriptions tailored for BLV individuals. Using this method, we have curated VideoA11y-40K, the largest and most comprehensive dataset of 40,000 videos described for BLV users. Rigorous experiments across 15 video categories, involving 347 sighted participants, 40 BLV participants, and seven professional describers, showed that VideoA11y descriptions outperform novice human annotations and are comparable to trained human annotations in clarity, accuracy, objectivity, descriptiveness, and user satisfaction. We evaluated models on VideoA11y-40K using both standard and custom metrics, demonstrating that MLLMs fine-tuned on this dataset produce high-quality accessible descriptions. Code and dataset are available at https://people-robots.github.io/VideoA11y.

We present VideoCLIP, a contrastive approach to pre-train a unified model for zero-shot video and text understanding, without using any labels on downstream tasks. VideoCLIP trains a transformer for video and text by contrasting temporally overlapping positive video-text pairs with hard negatives from nearest neighbor retrieval. Our experiments on a diverse series of downstream tasks, including sequence-level text-video retrieval, VideoQA, token-level action localization, and action segmentation reveal state-of-the-art performance, surpassing prior work, and in some cases even outperforming supervised approaches. Code is made available at https://github.com/pytorch/fairseq/tree/main/examples/MMPT.

While recent foundational video generators produce visually rich output, they still struggle with appearance drift, where objects gradually degrade or change inconsistently across frames, breaking visual coherence. We hypothesize that this is because there is no explicit supervision in terms of spatial tracking at the feature level. We propose Track4Gen, a spatially aware video generator that combines video diffusion loss with point tracking across frames, providing enhanced spatial supervision on the diffusion features. Track4Gen merges the video generation and point tracking tasks into a single network by making minimal changes to existing video generation architectures. Using Stable Video Diffusion as a backbone, Track4Gen demonstrates that it is possible to unify video generation and point tracking, which are typically handled as separate tasks. Our extensive evaluations show that Track4Gen effectively reduces appearance drift, resulting in temporally stable and visually coherent video generation. Project page: hyeonho99.github.io/track4gen

The robust association of the same objects across video frames in complex scenes is crucial for many applications, especially Multiple Object Tracking (MOT). Current methods predominantly rely on labeled domain-specific video datasets, which limits the cross-domain generalization of learned similarity embeddings. We propose MASA, a novel method for robust instance association learning, capable of matching any objects within videos across diverse domains without tracking labels. Leveraging the rich object segmentation from the Segment Anything Model (SAM), MASA learns instance-level correspondence through exhaustive data transformations. We treat the SAM outputs as dense object region proposals and learn to match those regions from a vast image collection. We further design a universal MASA adapter which can work in tandem with foundational segmentation or detection models and enable them to track any detected objects. Those combinations present strong zero-shot tracking ability in complex domains. Extensive tests on multiple challenging MOT and MOTS benchmarks indicate that the proposed method, using only unlabeled static images, achieves even better performance than state-of-the-art methods trained with fully annotated in-domain video sequences, in zero-shot association. Project Page: https://matchinganything.github.io/

We introduce ImmersePro, an innovative framework specifically designed to transform single-view videos into stereo videos. This framework utilizes a novel dual-branch architecture comprising a disparity branch and a context branch on video data by leveraging spatial-temporal attention mechanisms. ImmersePro employs implicit disparity guidance, enabling the generation of stereo pairs from video sequences without the need for explicit disparity maps, thus reducing potential errors associated with disparity estimation models. In addition to the technical advancements, we introduce the YouTube-SBS dataset, a comprehensive collection of 423 stereo videos sourced from YouTube. This dataset is unprecedented in its scale, featuring over 7 million stereo pairs, and is designed to facilitate training and benchmarking of stereo video generation models. Our experiments demonstrate the effectiveness of ImmersePro in producing high-quality stereo videos, offering significant improvements over existing methods. Compared to the best competitor stereo-from-mono we quantitatively improve the results by 11.76\% (L1), 6.39\% (SSIM), and 5.10\% (PSNR).

We are creating multimedia contents everyday and everywhere. While automatic content generation has played a fundamental challenge to multimedia community for decades, recent advances of deep learning have made this problem feasible. For example, the Generative Adversarial Networks (GANs) is a rewarding approach to synthesize images. Nevertheless, it is not trivial when capitalizing on GANs to generate videos. The difficulty originates from the intrinsic structure where a video is a sequence of visually coherent and semantically dependent frames. This motivates us to explore semantic and temporal coherence in designing GANs to generate videos. In this paper, we present a novel Temporal GANs conditioning on Captions, namely TGANs-C, in which the input to the generator network is a concatenation of a latent noise vector and caption embedding, and then is transformed into a frame sequence with 3D spatio-temporal convolutions. Unlike the naive discriminator which only judges pairs as fake or real, our discriminator additionally notes whether the video matches the correct caption. In particular, the discriminator network consists of three discriminators: video discriminator classifying realistic videos from generated ones and optimizes video-caption matching, frame discriminator discriminating between real and fake frames and aligning frames with the conditioning caption, and motion discriminator emphasizing the philosophy that the adjacent frames in the generated videos should be smoothly connected as in real ones. We qualitatively demonstrate the capability of our TGANs-C to generate plausible videos conditioning on the given captions on two synthetic datasets (SBMG and TBMG) and one real-world dataset (MSVD). Moreover, quantitative experiments on MSVD are performed to validate our proposal via Generative Adversarial Metric and human study.

Image matching is a fundamental computer vision problem. While learning-based methods achieve state-of-the-art performance on existing benchmarks, they generalize poorly to in-the-wild images. Such methods typically need to train separate models for different scene types and are impractical when the scene type is unknown in advance. One of the underlying problems is the limited scalability of existing data construction pipelines, which limits the diversity of standard image matching datasets. To address this problem, we propose GIM, a self-training framework for learning a single generalizable model based on any image matching architecture using internet videos, an abundant and diverse data source. Given an architecture, GIM first trains it on standard domain-specific datasets and then combines it with complementary matching methods to create dense labels on nearby frames of novel videos. These labels are filtered by robust fitting, and then enhanced by propagating them to distant frames. The final model is trained on propagated data with strong augmentations. We also propose ZEB, the first zero-shot evaluation benchmark for image matching. By mixing data from diverse domains, ZEB can thoroughly assess the cross-domain generalization performance of different methods. Applying GIM consistently improves the zero-shot performance of 3 state-of-the-art image matching architectures; with 50 hours of YouTube videos, the relative zero-shot performance improves by 8.4%-18.1%. GIM also enables generalization to extreme cross-domain data such as Bird Eye View (BEV) images of projected 3D point clouds (Fig. 1(c)). More importantly, our single zero-shot model consistently outperforms domain-specific baselines when evaluated on downstream tasks inherent to their respective domains. The video presentation is available at https://www.youtube.com/watch?v=FU_MJLD8LeY.

Methods for video motion prediction either estimate jointly the instantaneous motion of all points in a given video frame using optical flow or independently track the motion of individual points throughout the video. The latter is true even for powerful deep-learning methods that can track points through occlusions. Tracking points individually ignores the strong correlation that can exist between the points, for instance, because they belong to the same physical object, potentially harming performance. In this paper, we thus propose CoTracker, an architecture that jointly tracks multiple points throughout an entire video. This architecture combines several ideas from the optical flow and tracking literature in a new, flexible and powerful design. It is based on a transformer network that models the correlation of different points in time via specialised attention layers. The transformer iteratively updates an estimate of several trajectories. It can be applied in a sliding-window manner to very long videos, for which we engineer an unrolled training loop. It can track from one to several points jointly and supports adding new points to track at any time. The result is a flexible and powerful tracking algorithm that outperforms state-of-the-art methods in almost all benchmarks.

Video Highlight Detection and Moment Retrieval (HD/MR) are essential in video analysis. Recent joint prediction transformer models often overlook their cross-task dynamics and video-text alignment and refinement. Moreover, most models typically use limited, uni-directional attention mechanisms, resulting in weakly integrated representations and suboptimal performance in capturing the interdependence between video and text modalities. Although large-language and vision-language models (LLM/LVLMs) have gained prominence across various domains, their application in this field remains relatively underexplored. Here we propose VideoLights, a novel HD/MR framework addressing these limitations through (i) Convolutional Projection and Feature Refinement modules with an alignment loss for better video-text feature alignment, (ii) Bi-Directional Cross-Modal Fusion network for strongly coupled query-aware clip representations, and (iii) Uni-directional joint-task feedback mechanism enhancing both tasks through correlation. In addition, (iv) we introduce hard positive/negative losses for adaptive error penalization and improved learning, and (v) leverage LVLMs like BLIP-2 for enhanced multimodal feature integration and intelligent pretraining using synthetic data generated from LVLMs. Comprehensive experiments on QVHighlights, TVSum, and Charades-STA benchmarks demonstrate state-of-the-art performance. Codes and models are available at https://github.com/dpaul06/VideoLights .

Recently, much attention has been drawn to learning the underlying 3D structures of a scene from monocular videos in a fully self-supervised fashion. One of the most challenging aspects of this task is handling the independently moving objects as they break the rigid-scene assumption. For the first time, we show that pixel positional information can be exploited to learn SVDE (Single View Depth Estimation) from videos. Our proposed moving object (MO) masks, which are induced by shifted positional information (SPI) and referred to as `SPIMO' masks, are very robust and consistently remove the independently moving objects in the scenes, allowing for better learning of SVDE from videos. Additionally, we introduce a new adaptive quantization scheme that assigns the best per-pixel quantization curve for our depth discretization. Finally, we employ existing boosting techniques in a new way to further self-supervise the depth of the moving objects. With these features, our pipeline is robust against moving objects and generalizes well to high-resolution images, even when trained with small patches, yielding state-of-the-art (SOTA) results with almost 8.5x fewer parameters than the previous works that learn from videos. We present extensive experiments on KITTI and CityScapes that show the effectiveness of our method.

The foundation models have recently shown excellent performance on a variety of downstream tasks in computer vision. However, most existing vision foundation models simply focus on image-level pretraining and adpation, which are limited for dynamic and complex video-level understanding tasks. To fill the gap, we present general video foundation models, InternVideo, by taking advantage of both generative and discriminative self-supervised video learning. Specifically, InternVideo efficiently explores masked video modeling and video-language contrastive learning as the pretraining objectives, and selectively coordinates video representations of these two complementary frameworks in a learnable manner to boost various video applications. Without bells and whistles, InternVideo achieves state-of-the-art performance on 39 video datasets from extensive tasks including video action recognition/detection, video-language alignment, and open-world video applications. Especially, our methods can obtain 91.1% and 77.2% top-1 accuracy on the challenging Kinetics-400 and Something-Something V2 benchmarks, respectively. All of these results effectively show the generality of our InternVideo for video understanding. The code will be released at https://github.com/OpenGVLab/InternVideo .

Panoramic image stitching provides a unified, wide-angle view of a scene that extends beyond the camera's field of view. Stitching frames of a panning video into a panoramic photograph is a well-understood problem for stationary scenes, but when objects are moving, a still panorama cannot capture the scene. We present a method for synthesizing a panoramic video from a casually-captured panning video, as if the original video were captured with a wide-angle camera. We pose panorama synthesis as a space-time outpainting problem, where we aim to create a full panoramic video of the same length as the input video. Consistent completion of the space-time volume requires a powerful, realistic prior over video content and motion, for which we adapt generative video models. Existing generative models do not, however, immediately extend to panorama completion, as we show. We instead apply video generation as a component of our panorama synthesis system, and demonstrate how to exploit the strengths of the models while minimizing their limitations. Our system can create video panoramas for a range of in-the-wild scenes including people, vehicles, and flowing water, as well as stationary background features.

Generating videos from text is a challenging task due to its high computational requirements for training and infinite possible answers for evaluation. Existing works typically experiment on simple or small datasets, where the generalization ability is quite limited. In this work, we propose GODIVA, an open-domain text-to-video pretrained model that can generate videos from text in an auto-regressive manner using a three-dimensional sparse attention mechanism. We pretrain our model on Howto100M, a large-scale text-video dataset that contains more than 136 million text-video pairs. Experiments show that GODIVA not only can be fine-tuned on downstream video generation tasks, but also has a good zero-shot capability on unseen texts. We also propose a new metric called Relative Matching (RM) to automatically evaluate the video generation quality. Several challenges are listed and discussed as future work.

Recent works have successfully extended large-scale text-to-image models to the video domain, producing promising results but at a high computational cost and requiring a large amount of video data. In this work, we introduce ConditionVideo, a training-free approach to text-to-video generation based on the provided condition, video, and input text, by leveraging the power of off-the-shelf text-to-image generation methods (e.g., Stable Diffusion). ConditionVideo generates realistic dynamic videos from random noise or given scene videos. Our method explicitly disentangles the motion representation into condition-guided and scenery motion components. To this end, the ConditionVideo model is designed with a UNet branch and a control branch. To improve temporal coherence, we introduce sparse bi-directional spatial-temporal attention (sBiST-Attn). The 3D control network extends the conventional 2D controlnet model, aiming to strengthen conditional generation accuracy by additionally leveraging the bi-directional frames in the temporal domain. Our method exhibits superior performance in terms of frame consistency, clip score, and conditional accuracy, outperforming other compared methods.

Unsupervised learning of depth and ego-motion from unlabelled monocular videos has recently drawn great attention, which avoids the use of expensive ground truth in the supervised one. It achieves this by using the photometric errors between the target view and the synthesized views from its adjacent source views as the loss. Despite significant progress, the learning still suffers from occlusion and scene dynamics. This paper shows that carefully manipulating photometric errors can tackle these difficulties better. The primary improvement is achieved by a statistical technique that can mask out the invisible or nonstationary pixels in the photometric error map and thus prevents misleading the networks. With this outlier masking approach, the depth of objects moving in the opposite direction to the camera can be estimated more accurately. To the best of our knowledge, such scenarios have not been seriously considered in the previous works, even though they pose a higher risk in applications like autonomous driving. We also propose an efficient weighted multi-scale scheme to reduce the artifacts in the predicted depth maps. Extensive experiments on the KITTI dataset show the effectiveness of the proposed approaches. The overall system achieves state-of-theart performance on both depth and ego-motion estimation.

The vision and language generative models have been overgrown in recent years. For video generation, various open-sourced models and public-available services are released for generating high-visual quality videos. However, these methods often use a few academic metrics, for example, FVD or IS, to evaluate the performance. We argue that it is hard to judge the large conditional generative models from the simple metrics since these models are often trained on very large datasets with multi-aspect abilities. Thus, we propose a new framework and pipeline to exhaustively evaluate the performance of the generated videos. To achieve this, we first conduct a new prompt list for text-to-video generation by analyzing the real-world prompt list with the help of the large language model. Then, we evaluate the state-of-the-art video generative models on our carefully designed benchmarks, in terms of visual qualities, content qualities, motion qualities, and text-caption alignment with around 18 objective metrics. To obtain the final leaderboard of the models, we also fit a series of coefficients to align the objective metrics to the users' opinions. Based on the proposed opinion alignment method, our final score shows a higher correlation than simply averaging the metrics, showing the effectiveness of the proposed evaluation method.

The canonical approach to video captioning dictates a caption generation model to learn from offline-extracted dense video features. These feature extractors usually operate on video frames sampled at a fixed frame rate and are often trained on image/video understanding tasks, without adaption to video captioning data. In this work, we present SwinBERT, an end-to-end transformer-based model for video captioning, which takes video frame patches directly as inputs, and outputs a natural language description. Instead of leveraging multiple 2D/3D feature extractors, our method adopts a video transformer to encode spatial-temporal representations that can adapt to variable lengths of video input without dedicated design for different frame rates. Based on this model architecture, we show that video captioning can benefit significantly from more densely sampled video frames as opposed to previous successes with sparsely sampled video frames for video-and-language understanding tasks (e.g., video question answering). Moreover, to avoid the inherent redundancy in consecutive video frames, we propose adaptively learning a sparse attention mask and optimizing it for task-specific performance improvement through better long-range video sequence modeling. Through extensive experiments on 5 video captioning datasets, we show that SwinBERT achieves across-the-board performance improvements over previous methods, often by a large margin. The learned sparse attention masks in addition push the limit to new state of the arts, and can be transferred between different video lengths and between different datasets. Code is available at https://github.com/microsoft/SwinBERT

We propose a self-supervised method for learning motion-focused video representations. Existing approaches minimize distances between temporally augmented videos, which maintain high spatial similarity. We instead propose to learn similarities between videos with identical local motion dynamics but an otherwise different appearance. We do so by adding synthetic motion trajectories to videos which we refer to as tubelets. By simulating different tubelet motions and applying transformations, such as scaling and rotation, we introduce motion patterns beyond what is present in the pretraining data. This allows us to learn a video representation that is remarkably data-efficient: our approach maintains performance when using only 25% of the pretraining videos. Experiments on 10 diverse downstream settings demonstrate our competitive performance and generalizability to new domains and fine-grained actions.

Understanding human actions from videos of first-person view poses significant challenges. Most prior approaches explore representation learning on egocentric videos only, while overlooking the potential benefit of exploiting existing large-scale third-person videos. In this paper, (1) we develop EgoInstructor, a retrieval-augmented multimodal captioning model that automatically retrieves semantically relevant third-person instructional videos to enhance the video captioning of egocentric videos. (2) For training the cross-view retrieval module, we devise an automatic pipeline to discover ego-exo video pairs from distinct large-scale egocentric and exocentric datasets. (3) We train the cross-view retrieval module with a novel EgoExoNCE loss that pulls egocentric and exocentric video features closer by aligning them to shared text features that describe similar actions. (4) Through extensive experiments, our cross-view retrieval module demonstrates superior performance across seven benchmarks. Regarding egocentric video captioning, EgoInstructor exhibits significant improvements by leveraging third-person videos as references.

We introduce InternVideo2, a new video foundation model (ViFM) that achieves the state-of-the-art performance in action recognition, video-text tasks, and video-centric dialogue. Our approach employs a progressive training paradigm that unifies the different self- or weakly-supervised learning frameworks of masked video token reconstruction, cross-modal contrastive learning, and next token prediction. Different training stages would guide our model to capture different levels of structure and semantic information through different pretext tasks. At the data level, we prioritize the spatiotemporal consistency by semantically segmenting videos and generating video-audio-speech captions. This improves the alignment between video and text. We scale both data and model size for our InternVideo2. Through extensive experiments, we validate our designs and demonstrate the state-of-the-art performance on over 60 video and audio tasks. Notably, our model outperforms others on various video-related captioning, dialogue, and long video understanding benchmarks, highlighting its ability to reason and comprehend long temporal contexts. Code and models are available at https://github.com/OpenGVLab/InternVideo2/.

Long video understanding is a significant and ongoing challenge in the intersection of multimedia and artificial intelligence. Employing large language models (LLMs) for comprehending video becomes an emerging and promising method. However, this approach incurs high computational costs due to the extensive array of video tokens, experiences reduced visual clarity as a consequence of token aggregation, and confronts challenges arising from irrelevant visual tokens while answering video-related questions. To alleviate these issues, we present an Interactive Visual Adapter (IVA) within LLMs, designed to enhance interaction with fine-grained visual elements. Specifically, we first transform long videos into temporal video tokens via leveraging a visual encoder alongside a pretrained causal transformer, then feed them into LLMs with the video instructions. Subsequently, we integrated IVA, which contains a lightweight temporal frame selector and a spatial feature interactor, within the internal blocks of LLMs to capture instruction-aware and fine-grained visual signals. Consequently, the proposed video-LLM facilitates a comprehensive understanding of long video content through appropriate long video modeling and precise visual interactions. We conducted extensive experiments on nine video understanding benchmarks and experimental results show that our interactive visual adapter significantly improves the performance of video LLMs on long video QA tasks. Ablation studies further verify the effectiveness of IVA in long and short video understandings.

Recently, there is a surge in interest surrounding video large language models (Video LLMs). However, existing benchmarks fail to provide a comprehensive feedback on the temporal perception ability of Video LLMs. On the one hand, most of them are unable to distinguish between different temporal aspects (e.g., speed, direction) and thus cannot reflect the nuanced performance on these specific aspects. On the other hand, they are limited in the diversity of task formats (e.g., only multi-choice QA), which hinders the understanding of how temporal perception performance may vary across different types of tasks. Motivated by these two problems, we propose the TempCompass benchmark, which introduces a diversity of temporal aspects and task formats. To collect high-quality test data, we devise two novel strategies: (1) In video collection, we construct conflicting videos that share the same static content but differ in a specific temporal aspect, which prevents Video LLMs from leveraging single-frame bias or language priors. (2) To collect the task instructions, we propose a paradigm where humans first annotate meta-information for a video and then an LLM generates the instruction. We also design an LLM-based approach to automatically and accurately evaluate the responses from Video LLMs. Based on TempCompass, we comprehensively evaluate 8 state-of-the-art (SOTA) Video LLMs and 3 Image LLMs, and reveal the discerning fact that these models exhibit notably poor temporal perception ability. The data and evaluation code are available at https://github.com/llyx97/TempCompass.

Recently, advancements in video synthesis have attracted significant attention. Video synthesis models such as AnimateDiff and Stable Video Diffusion have demonstrated the practical applicability of diffusion models in creating dynamic visual content. The emergence of SORA has further spotlighted the potential of video generation technologies. Nonetheless, the extension of video lengths has been constrained by the limitations in computational resources. Most existing video synthesis models can only generate short video clips. In this paper, we propose a novel post-tuning methodology for video synthesis models, called ExVideo. This approach is designed to enhance the capability of current video synthesis models, allowing them to produce content over extended temporal durations while incurring lower training expenditures. In particular, we design extension strategies across common temporal model architectures respectively, including 3D convolution, temporal attention, and positional embedding. To evaluate the efficacy of our proposed post-tuning approach, we conduct extension training on the Stable Video Diffusion model. Our approach augments the model's capacity to generate up to 5times its original number of frames, requiring only 1.5k GPU hours of training on a dataset comprising 40k videos. Importantly, the substantial increase in video length doesn't compromise the model's innate generalization capabilities, and the model showcases its advantages in generating videos of diverse styles and resolutions. We will release the source code and the enhanced model publicly.

We present the ShareGPT4Video series, aiming to facilitate the video understanding of large video-language models (LVLMs) and the video generation of text-to-video models (T2VMs) via dense and precise captions. The series comprises: 1) ShareGPT4Video, 40K GPT4V annotated dense captions of videos with various lengths and sources, developed through carefully designed data filtering and annotating strategy. 2) ShareCaptioner-Video, an efficient and capable captioning model for arbitrary videos, with 4.8M high-quality aesthetic videos annotated by it. 3) ShareGPT4Video-8B, a simple yet superb LVLM that reached SOTA performance on three advancing video benchmarks. To achieve this, taking aside the non-scalable costly human annotators, we find using GPT4V to caption video with a naive multi-frame or frame-concatenation input strategy leads to less detailed and sometimes temporal-confused results. We argue the challenge of designing a high-quality video captioning strategy lies in three aspects: 1) Inter-frame precise temporal change understanding. 2) Intra-frame detailed content description. 3) Frame-number scalability for arbitrary-length videos. To this end, we meticulously designed a differential video captioning strategy, which is stable, scalable, and efficient for generating captions for videos with arbitrary resolution, aspect ratios, and length. Based on it, we construct ShareGPT4Video, which contains 40K high-quality videos spanning a wide range of categories, and the resulting captions encompass rich world knowledge, object attributes, camera movements, and crucially, detailed and precise temporal descriptions of events. Based on ShareGPT4Video, we further develop ShareCaptioner-Video, a superior captioner capable of efficiently generating high-quality captions for arbitrary videos...

Training a Neural Radiance Field (NeRF) without pre-computed camera poses is challenging. Recent advances in this direction demonstrate the possibility of jointly optimising a NeRF and camera poses in forward-facing scenes. However, these methods still face difficulties during dramatic camera movement. We tackle this challenging problem by incorporating undistorted monocular depth priors. These priors are generated by correcting scale and shift parameters during training, with which we are then able to constrain the relative poses between consecutive frames. This constraint is achieved using our proposed novel loss functions. Experiments on real-world indoor and outdoor scenes show that our method can handle challenging camera trajectories and outperforms existing methods in terms of novel view rendering quality and pose estimation accuracy. Our project page is https://nope-nerf.active.vision.

This work aims to learn a high-quality text-to-video (T2V) generative model by leveraging a pre-trained text-to-image (T2I) model as a basis. It is a highly desirable yet challenging task to simultaneously a) accomplish the synthesis of visually realistic and temporally coherent videos while b) preserving the strong creative generation nature of the pre-trained T2I model. To this end, we propose LaVie, an integrated video generation framework that operates on cascaded video latent diffusion models, comprising a base T2V model, a temporal interpolation model, and a video super-resolution model. Our key insights are two-fold: 1) We reveal that the incorporation of simple temporal self-attentions, coupled with rotary positional encoding, adequately captures the temporal correlations inherent in video data. 2) Additionally, we validate that the process of joint image-video fine-tuning plays a pivotal role in producing high-quality and creative outcomes. To enhance the performance of LaVie, we contribute a comprehensive and diverse video dataset named Vimeo25M, consisting of 25 million text-video pairs that prioritize quality, diversity, and aesthetic appeal. Extensive experiments demonstrate that LaVie achieves state-of-the-art performance both quantitatively and qualitatively. Furthermore, we showcase the versatility of pre-trained LaVie models in various long video generation and personalized video synthesis applications.

Due to the resource-intensive nature of training vision-language models on expansive video data, a majority of studies have centered on adapting pre-trained image-language models to the video domain. Dominant pipelines propose to tackle the visual discrepancies with additional temporal learners while overlooking the substantial discrepancy for web-scaled descriptive narratives and concise action category names, leading to less distinct semantic space and potential performance limitations. In this work, we prioritize the refinement of text knowledge to facilitate generalizable video recognition. To address the limitations of the less distinct semantic space of category names, we prompt a large language model (LLM) to augment action class names into Spatio-Temporal Descriptors thus bridging the textual discrepancy and serving as a knowledge base for general recognition. Moreover, to assign the best descriptors with different video instances, we propose Optimal Descriptor Solver, forming the video recognition problem as solving the optimal matching flow across frame-level representations and descriptors. Comprehensive evaluations in zero-shot, few-shot, and fully supervised video recognition highlight the effectiveness of our approach. Our best model achieves a state-of-the-art zero-shot accuracy of 75.1% on Kinetics-600.

The goal of spatial-temporal action detection is to determine the time and place where each person's action occurs in a video and classify the corresponding action category. Most of the existing methods adopt fully-supervised learning, which requires a large amount of training data, making it very difficult to achieve zero-shot learning. In this paper, we propose to utilize a pre-trained visual-language model to extract the representative image and text features, and model the relationship between these features through different interaction modules to obtain the interaction feature. In addition, we use this feature to prompt each label to obtain more appropriate text features. Finally, we calculate the similarity between the interaction feature and the text feature for each label to determine the action category. Our experiments on J-HMDB and UCF101-24 datasets demonstrate that the proposed interaction module and prompting make the visual-language features better aligned, thus achieving excellent accuracy for zero-shot spatio-temporal action detection. The code will be available at https://github.com/webber2933/iCLIP.

In this paper our objectives are, first, networks that can embed audio and visual inputs into a common space that is suitable for cross-modal retrieval; and second, a network that can localize the object that sounds in an image, given the audio signal. We achieve both these objectives by training from unlabelled video using only audio-visual correspondence (AVC) as the objective function. This is a form of cross-modal self-supervision from video. To this end, we design new network architectures that can be trained for cross-modal retrieval and localizing the sound source in an image, by using the AVC task. We make the following contributions: (i) show that audio and visual embeddings can be learnt that enable both within-mode (e.g. audio-to-audio) and between-mode retrieval; (ii) explore various architectures for the AVC task, including those for the visual stream that ingest a single image, or multiple images, or a single image and multi-frame optical flow; (iii) show that the semantic object that sounds within an image can be localized (using only the sound, no motion or flow information); and (iv) give a cautionary tale on how to avoid undesirable shortcuts in the data preparation.

Tracking dense 3D motion from monocular videos remains challenging, particularly when aiming for pixel-level precision over long sequences. We introduce \Approach, a novel method that efficiently tracks every pixel in 3D space, enabling accurate motion estimation across entire videos. Our approach leverages a joint global-local attention mechanism for reduced-resolution tracking, followed by a transformer-based upsampler to achieve high-resolution predictions. Unlike existing methods, which are limited by computational inefficiency or sparse tracking, \Approach delivers dense 3D tracking at scale, running over 8x faster than previous methods while achieving state-of-the-art accuracy. Furthermore, we explore the impact of depth representation on tracking performance and identify log-depth as the optimal choice. Extensive experiments demonstrate the superiority of \Approach on multiple benchmarks, achieving new state-of-the-art results in both 2D and 3D dense tracking tasks. Our method provides a robust solution for applications requiring fine-grained, long-term motion tracking in 3D space.

Motion, scene and object are three primary visual components of a video. In particular, objects represent the foreground, scenes represent the background, and motion traces their dynamics. Based on this insight, we propose a two-stage MOtion, Scene and Object decomposition framework (MOSO) for video prediction, consisting of MOSO-VQVAE and MOSO-Transformer. In the first stage, MOSO-VQVAE decomposes a previous video clip into the motion, scene and object components, and represents them as distinct groups of discrete tokens. Then, in the second stage, MOSO-Transformer predicts the object and scene tokens of the subsequent video clip based on the previous tokens and adds dynamic motion at the token level to the generated object and scene tokens. Our framework can be easily extended to unconditional video generation and video frame interpolation tasks. Experimental results demonstrate that our method achieves new state-of-the-art performance on five challenging benchmarks for video prediction and unconditional video generation: BAIR, RoboNet, KTH, KITTI and UCF101. In addition, MOSO can produce realistic videos by combining objects and scenes from different videos.

Significant advancements in video diffusion models have brought substantial progress to the field of text-to-video (T2V) synthesis. However, existing T2V synthesis model struggle to accurately generate complex motion dynamics, leading to a reduction in video realism. One possible solution is to collect massive data and train the model on it, but this would be extremely expensive. To alleviate this problem, in this paper, we reformulate the typical T2V generation process as a search-based generation pipeline. Instead of scaling up the model training, we employ existing videos as the motion prior database. Specifically, we divide T2V generation process into two steps: (i) For a given prompt input, we search existing text-video datasets to find videos with text labels that closely match the prompt motions. We propose a tailored search algorithm that emphasizes object motion features. (ii) Retrieved videos are processed and distilled into motion priors to fine-tune a pre-trained base T2V model, followed by generating desired videos using input prompt. By utilizing the priors gleaned from the searched videos, we enhance the realism of the generated videos' motion. All operations can be finished on a single NVIDIA RTX 4090 GPU. We validate our method against state-of-the-art T2V models across diverse prompt inputs. The code will be public.

Learning discriminative spatiotemporal representation is the key problem of video understanding. Recently, Vision Transformers (ViTs) have shown their power in learning long-term video dependency with self-attention. Unfortunately, they exhibit limitations in tackling local video redundancy, due to the blind global comparison among tokens. UniFormer has successfully alleviated this issue, by unifying convolution and self-attention as a relation aggregator in the transformer format. However, this model has to require a tiresome and complicated image-pretraining phrase, before being finetuned on videos. This blocks its wide usage in practice. On the contrary, open-sourced ViTs are readily available and well-pretrained with rich image supervision. Based on these observations, we propose a generic paradigm to build a powerful family of video networks, by arming the pretrained ViTs with efficient UniFormer designs. We call this family UniFormerV2, since it inherits the concise style of the UniFormer block. But it contains brand-new local and global relation aggregators, which allow for preferable accuracy-computation balance by seamlessly integrating advantages from both ViTs and UniFormer. Without any bells and whistles, our UniFormerV2 gets the state-of-the-art recognition performance on 8 popular video benchmarks, including scene-related Kinetics-400/600/700 and Moments in Time, temporal-related Something-Something V1/V2, untrimmed ActivityNet and HACS. In particular, it is the first model to achieve 90% top-1 accuracy on Kinetics-400, to our best knowledge. Code will be available at https://github.com/OpenGVLab/UniFormerV2.

We propose a self-supervised method to learn feature representations from videos. A standard approach in traditional self-supervised methods uses positive-negative data pairs to train with contrastive learning strategy. In such a case, different modalities of the same video are treated as positives and video clips from a different video are treated as negatives. Because the spatio-temporal information is important for video representation, we extend the negative samples by introducing intra-negative samples, which are transformed from the same anchor video by breaking temporal relations in video clips. With the proposed Inter-Intra Contrastive (IIC) framework, we can train spatio-temporal convolutional networks to learn video representations. There are many flexible options in our IIC framework and we conduct experiments by using several different configurations. Evaluations are conducted on video retrieval and video recognition tasks using the learned video representation. Our proposed IIC outperforms current state-of-the-art results by a large margin, such as 16.7% and 9.5% points improvements in top-1 accuracy on UCF101 and HMDB51 datasets for video retrieval, respectively. For video recognition, improvements can also be obtained on these two benchmark datasets. Code is available at https://github.com/BestJuly/Inter-intra-video-contrastive-learning.

Despite an exciting new wave of multimodal machine learning models, current approaches still struggle to interpret the complex contextual relationships between the different modalities present in videos. Going beyond existing methods that emphasize simple activities or objects, we propose a new model-agnostic approach for generating detailed textual descriptions that captures multimodal video information. Our method leverages the extensive knowledge learnt by large language models, such as GPT-3.5 or Llama2, to reason about textual descriptions of the visual and aural modalities, obtained from BLIP-2, Whisper and ImageBind. Without needing additional finetuning of video-text models or datasets, we demonstrate that available LLMs have the ability to use these multimodal textual descriptions as proxies for ``sight'' or ``hearing'' and perform zero-shot multimodal classification of videos in-context. Our evaluations on popular action recognition benchmarks, such as UCF-101 or Kinetics, show these context-rich descriptions can be successfully used in video understanding tasks. This method points towards a promising new research direction in multimodal classification, demonstrating how an interplay between textual, visual and auditory machine learning models can enable more holistic video understanding.

In this paper, we propose VideoLLaMA3, a more advanced multimodal foundation model for image and video understanding. The core design philosophy of VideoLLaMA3 is vision-centric. The meaning of "vision-centric" is two-fold: the vision-centric training paradigm and vision-centric framework design. The key insight of our vision-centric training paradigm is that high-quality image-text data is crucial for both image and video understanding. Instead of preparing massive video-text datasets, we focus on constructing large-scale and high-quality image-text datasets. VideoLLaMA3 has four training stages: 1) vision-centric alignment stage, which warms up the vision encoder and projector; 2) vision-language pretraining stage, which jointly tunes the vision encoder, projector, and LLM with large-scale image-text data covering multiple types (including scene images, documents, charts) as well as text-only data. 3) multi-task fine-tuning stage, which incorporates image-text SFT data for downstream tasks and video-text data to establish a foundation for video understanding. 4) video-centric fine-tuning, which further improves the model's capability in video understanding. As for the framework design, to better capture fine-grained details in images, the pretrained vision encoder is adapted to encode images of varying sizes into vision tokens with corresponding numbers, rather than a fixed number of tokens. For video inputs, we reduce the number of vision tokens according to their similarity so that the representation of videos will be more precise and compact. Benefit from vision-centric designs, VideoLLaMA3 achieves compelling performances in both image and video understanding benchmarks.

Several unsupervised and self-supervised approaches have been developed in recent years to learn visual features from large-scale unlabeled datasets. Their main drawback however is that these methods are hardly able to recognize visual features of the same object if it is simply rotated or the perspective of the camera changes. To overcome this limitation and at the same time exploit a useful source of supervision, we take into account video object tracks. Following the intuition that two patches in a track should have similar visual representations in a learned feature space, we adopt an unsupervised clustering-based approach and constrain such representations to be labeled as the same category since they likely belong to the same object or object part. Experimental results on two downstream tasks on different datasets demonstrate the effectiveness of our Online Deep Clustering with Video Track Consistency (ODCT) approach compared to prior work, which did not leverage temporal information. In addition we show that exploiting an unsupervised class-agnostic, yet noisy, track generator yields to better accuracy compared to relying on costly and precise track annotations.

Most current LLM-based models for video understanding can process videos within minutes. However, they struggle with lengthy videos due to challenges such as "noise and redundancy", as well as "memory and computation" constraints. In this paper, we present Goldfish, a methodology tailored for comprehending videos of arbitrary lengths. We also introduce the TVQA-long benchmark, specifically designed to evaluate models' capabilities in understanding long videos with questions in both vision and text content. Goldfish approaches these challenges with an efficient retrieval mechanism that initially gathers the top-k video clips relevant to the instruction before proceeding to provide the desired response. This design of the retrieval mechanism enables the Goldfish to efficiently process arbitrarily long video sequences, facilitating its application in contexts such as movies or television series. To facilitate the retrieval process, we developed MiniGPT4-Video that generates detailed descriptions for the video clips. In addressing the scarcity of benchmarks for long video evaluation, we adapted the TVQA short video benchmark for extended content analysis by aggregating questions from entire episodes, thereby shifting the evaluation from partial to full episode comprehension. We attained a 41.78% accuracy rate on the TVQA-long benchmark, surpassing previous methods by 14.94%. Our MiniGPT4-Video also shows exceptional performance in short video comprehension, exceeding existing state-of-the-art methods by 3.23%, 2.03%, 16.5% and 23.59% on the MSVD, MSRVTT, TGIF, and TVQA short video benchmarks, respectively. These results indicate that our models have significant improvements in both long and short-video understanding. Our models and code have been made publicly available at https://vision-cair.github.io/Goldfish_website/

We propose a self-supervised algorithm to learn representations from egocentric video data. Recently, significant efforts have been made to capture humans interacting with their own environments as they go about their daily activities. In result, several large egocentric datasets of interaction-rich multi-modal data have emerged. However, learning representations from videos can be challenging. First, given the uncurated nature of long-form continuous videos, learning effective representations require focusing on moments in time when interactions take place. Second, visual representations of daily activities should be sensitive to changes in the state of the environment. However, current successful multi-modal learning frameworks encourage representation invariance over time. To address these challenges, we leverage audio signals to identify moments of likely interactions which are conducive to better learning. We also propose a novel self-supervised objective that learns from audible state changes caused by interactions. We validate these contributions extensively on two large-scale egocentric datasets, EPIC-Kitchens-100 and the recently released Ego4D, and show improvements on several downstream tasks, including action recognition, long-term action anticipation, and object state change classification.

Recent advances in multimodal Large Language Models (LLMs) have shown great success in understanding multi-modal contents. For video understanding tasks, training-based video LLMs are difficult to build due to the scarcity of high-quality, curated video-text paired data. In contrast, paired image-text data are much easier to obtain, and there is substantial similarity between images and videos. Consequently, extending image LLMs for video understanding tasks presents an appealing alternative. Developing effective strategies for compressing visual tokens from multiple frames is a promising way to leverage the powerful pre-trained image LLM. In this work, we explore the limitations of the existing compression strategies for building a training-free video LLM. The findings lead to our method TS-LLaVA, which constructs visual tokens through a Thumbnail-and-Sampling strategy. Given a video, we select few equidistant frames from all input frames to construct a Thumbnail image as a detailed visual cue, complemented by Sampled visual tokens from all input frames. Our method establishes the new state-of-the-art performance among training-free video LLMs on various benchmarks. Notably, our 34B model outperforms GPT-4V on the MVBench benchmark, and achieves performance comparable to the 72B training-based video LLM, Video-LLaMA2, on the challenging MLVU benchmark. Code is available at https://github.com/tingyu215/TS-LLaVA.

Video depth estimation lifts monocular video clips to 3D by inferring dense depth at every frame. Recent advances in single-image depth estimation, brought about by the rise of large foundation models and the use of synthetic training data, have fueled a renewed interest in video depth. However, naively applying a single-image depth estimator to every frame of a video disregards temporal continuity, which not only leads to flickering but may also break when camera motion causes sudden changes in depth range. An obvious and principled solution would be to build on top of video foundation models, but these come with their own limitations; including expensive training and inference, imperfect 3D consistency, and stitching routines for the fixed-length (short) outputs. We take a step back and demonstrate how to turn a single-image latent diffusion model (LDM) into a state-of-the-art video depth estimator. Our model, which we call RollingDepth, has two main ingredients: (i) a multi-frame depth estimator that is derived from a single-image LDM and maps very short video snippets (typically frame triplets) to depth snippets. (ii) a robust, optimization-based registration algorithm that optimally assembles depth snippets sampled at various different frame rates back into a consistent video. RollingDepth is able to efficiently handle long videos with hundreds of frames and delivers more accurate depth videos than both dedicated video depth estimators and high-performing single-frame models. Project page: rollingdepth.github.io.

Recent years have seen substantial progress in diffusion-based controllable video generation. However, achieving precise control in complex scenarios, including fine-grained object parts, sophisticated motion trajectories, and coherent background movement, remains a challenge. In this paper, we introduce TrackGo, a novel approach that leverages free-form masks and arrows for conditional video generation. This method offers users with a flexible and precise mechanism for manipulating video content. We also propose the TrackAdapter for control implementation, an efficient and lightweight adapter designed to be seamlessly integrated into the temporal self-attention layers of a pretrained video generation model. This design leverages our observation that the attention map of these layers can accurately activate regions corresponding to motion in videos. Our experimental results demonstrate that our new approach, enhanced by the TrackAdapter, achieves state-of-the-art performance on key metrics such as FVD, FID, and ObjMC scores. The project page of TrackGo can be found at: https://zhtjtcz.github.io/TrackGo-Page/

Efficiently retrieving and synthesizing information from large-scale multimodal collections has become a critical challenge. However, existing video retrieval datasets suffer from scope limitations, primarily focusing on matching descriptive but vague queries with small collections of professionally edited, English-centric videos. To address this gap, we introduce MultiVENT 2.0, a large-scale, multilingual event-centric video retrieval benchmark featuring a collection of more than 218,000 news videos and 3,906 queries targeting specific world events. These queries specifically target information found in the visual content, audio, embedded text, and text metadata of the videos, requiring systems leverage all these sources to succeed at the task. Preliminary results show that state-of-the-art vision-language models struggle significantly with this task, and while alternative approaches show promise, they are still insufficient to adequately address this problem. These findings underscore the need for more robust multimodal retrieval systems, as effective video retrieval is a crucial step towards multimodal content understanding and generation tasks.

Video-Language Pretraining (VLP), which aims to learn transferable representation to advance a wide range of video-text downstream tasks, has recently received increasing attention. Best performing works rely on large-scale, 3rd-person video-text datasets, such as HowTo100M. In this work, we exploit the recently released Ego4D dataset to pioneer Egocentric VLP along three directions. (i) We create EgoClip, a 1st-person video-text pretraining dataset comprising 3.8M clip-text pairs well-chosen from Ego4D, covering a large variety of human daily activities. (ii) We propose a novel pretraining objective, dubbed EgoNCE, which adapts video-text contrastive learning to the egocentric domain by mining egocentric-aware positive and negative samples. (iii) We introduce EgoMCQ, a development benchmark that is close to EgoClip and hence can support effective validation and fast exploration of our design decisions in EgoClip and EgoNCE. Furthermore, we demonstrate strong performance on five egocentric downstream tasks across three datasets: video-text retrieval on EPIC-KITCHENS-100; action recognition on Charades-Ego; natural language query, moment query, and object state change classification on Ego4D challenge benchmarks. The dataset and code are available at https://github.com/showlab/EgoVLP.

View-predictive generative models provide strong priors for lifting object-centric images and videos into 3D and 4D through rendering and score distillation objectives. A question then remains: what about lifting complete multi-object dynamic scenes? There are two challenges in this direction: First, rendering error gradients are often insufficient to recover fast object motion, and second, view predictive generative models work much better for objects than whole scenes, so, score distillation objectives cannot currently be applied at the scene level directly. We present DreamScene4D, the first approach to generate 3D dynamic scenes of multiple objects from monocular videos via 360-degree novel view synthesis. Our key insight is a "decompose-recompose" approach that factorizes the video scene into the background and object tracks, while also factorizing object motion into 3 components: object-centric deformation, object-to-world-frame transformation, and camera motion. Such decomposition permits rendering error gradients and object view-predictive models to recover object 3D completions and deformations while bounding box tracks guide the large object movements in the scene. We show extensive results on challenging DAVIS, Kubric, and self-captured videos with quantitative comparisons and a user preference study. Besides 4D scene generation, DreamScene4D obtains accurate 2D persistent point track by projecting the inferred 3D trajectories to 2D. We will release our code and hope our work will stimulate more research on fine-grained 4D understanding from videos.

This paper investigates the challenge of extracting highlight moments from videos. To perform this task, a system needs to understand what constitutes a highlight for arbitrary video domains while at the same time being able to scale across different domains. Our key insight is that photographs taken by photographers tend to capture the most remarkable or photogenic moments of an activity. Drawing on this insight, we present Videogenic, a system capable of creating domain-specific highlight videos for a wide range of domains. In a human evaluation study (N=50), we show that a high-quality photograph collection combined with CLIP-based retrieval (which uses a neural network with semantic knowledge of images) can serve as an excellent prior for finding video highlights. In a within-subjects expert study (N=12), we demonstrate the usefulness of Videogenic in helping video editors create highlight videos with lighter workload, shorter task completion time, and better usability.

Scaling has not yet been convincingly demonstrated for pure self-supervised learning from video. However, prior work has focused evaluations on semantic-related tasks x2013 action classification, ImageNet classification, etc. In this paper we focus on evaluating self-supervised learning on non-semantic vision tasks that are more spatial (3D) and temporal (+1D = 4D), such as camera pose estimation, point and object tracking, and depth estimation. We show that by learning from very large video datasets, masked auto-encoding (MAE) with transformer video models actually scales, consistently improving performance on these 4D tasks, as model size increases from 20M all the way to the largest by far reported self-supervised video model x2013 22B parameters. Rigorous apples-to-apples comparison with many recent image and video models demonstrates the benefits of scaling 4D representations.

What makes good video representations for video understanding, such as anticipating future activities, or answering video-conditioned questions? While earlier approaches focus on end-to-end learning directly from video pixels, we propose to revisit text-based representations, such as discrete action labels, or free-form video captions, which are interpretable and can be directly consumed by large language models (LLMs). Intuitively, different video understanding tasks may require representations that are complementary and at different granularities. To this end, we propose versatile action models (Vamos), a learning framework powered by a large language model as the "reasoner", and can flexibly leverage visual embeddings, action labels, and free-form descriptions extracted from videos as its input. We evaluate Vamos on four complementary video understanding benchmarks, Ego4D, Next-QA, IntentQA, and EgoSchema, on its capability to model temporal dynamics, encode visual history, and perform reasoning. Surprisingly, we observe that text-based representations consistently achieve competitive performance on all benchmarks, and that visual embeddings provide marginal or no performance improvement, demonstrating the effectiveness of text-based video representation in the LLM era. We perform extensive ablation study and qualitative analysis to support our observations, and achieve state-of-the-art performance on three benchmarks.

Recent advances in vision-language models have significantly propelled video understanding. Existing datasets and tasks, however, have notable limitations. Most datasets are confined to short videos with limited events and narrow narratives. For example, datasets with instructional and egocentric videos often document the activities of one person in a single scene. Although some movie datasets offer richer content, they are often limited to short-term tasks, lack publicly available videos and frequently encounter data leakage given the use of movie forums and other resources in LLM training. To address the above limitations, we propose the Short Film Dataset (SFD) with 1,078 publicly available amateur movies, a wide variety of genres and minimal data leakage issues. SFD offers long-term story-oriented video tasks in the form of multiple-choice and open-ended question answering. Our extensive experiments emphasize the need for long-term reasoning to solve SFD tasks. Notably, we find strong signals in movie transcripts leading to the on-par performance of people and LLMs. We also show significantly lower performance of current models compared to people when using vision data alone.

Diffusion models have demonstrated great success in text-to-video (T2V) generation. However, existing methods may face challenges when handling complex (long) video generation scenarios that involve multiple objects or dynamic changes in object numbers. To address these limitations, we propose VideoTetris, a novel framework that enables compositional T2V generation. Specifically, we propose spatio-temporal compositional diffusion to precisely follow complex textual semantics by manipulating and composing the attention maps of denoising networks spatially and temporally. Moreover, we propose an enhanced video data preprocessing to enhance the training data regarding motion dynamics and prompt understanding, equipped with a new reference frame attention mechanism to improve the consistency of auto-regressive video generation. Extensive experiments demonstrate that our VideoTetris achieves impressive qualitative and quantitative results in compositional T2V generation. Code is available at: https://github.com/YangLing0818/VideoTetris

We study unsupervised video representation learning that seeks to learn both motion and appearance features from unlabeled video only, which can be reused for downstream tasks such as action recognition. This task, however, is extremely challenging due to 1) the highly complex spatial-temporal information in videos; and 2) the lack of labeled data for training. Unlike the representation learning for static images, it is difficult to construct a suitable self-supervised task to well model both motion and appearance features. More recently, several attempts have been made to learn video representation through video playback speed prediction. However, it is non-trivial to obtain precise speed labels for the videos. More critically, the learnt models may tend to focus on motion pattern and thus may not learn appearance features well. In this paper, we observe that the relative playback speed is more consistent with motion pattern, and thus provide more effective and stable supervision for representation learning. Therefore, we propose a new way to perceive the playback speed and exploit the relative speed between two video clips as labels. In this way, we are able to well perceive speed and learn better motion features. Moreover, to ensure the learning of appearance features, we further propose an appearance-focused task, where we enforce the model to perceive the appearance difference between two video clips. We show that optimizing the two tasks jointly consistently improves the performance on two downstream tasks, namely action recognition and video retrieval. Remarkably, for action recognition on UCF101 dataset, we achieve 93.7% accuracy without the use of labeled data for pre-training, which outperforms the ImageNet supervised pre-trained model. Code and pre-trained models can be found at https://github.com/PeihaoChen/RSPNet.