File size: 2,030 Bytes
fce460f
 
 
 
4edeca4
 
 
 
 
 
fce460f
 
4edeca4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
---
tags:
- model_hub_mixin
- pytorch_model_hub_mixin
language:
- en
base_model:
- google/vit-base-patch16-224
- google-bert/bert-base-uncased
- MCG-NJU/videomae-base
---

# πŸ–ΌοΈπŸ“ OneEncoder: A Unified Text & Image & Video Model  

**OneEncoder** is a lightweight framework for cross-modal alignment, focusing on efficiently integrating **text, image and video** (with future extensions to other modalities). Unlike traditional methods relying on massive modality-specific encoders, OneEncoder progressively aligns different data types, making it cost-effective and performant even on small paired datasets.  

## πŸš€ Key Features  
βœ… **Multimodal Alignment**: Initially supports **text & image & video**, with extension to other modalities.  
βœ… **Lightweight & Efficient**: Avoids full retraining when adding new modalities.  
βœ… **Superior Performance**: Outperforms models that require large specialized datasets.  

## 🎯 Applications  
- **Visual Question Answering (VQA)**  
- **Image-Text-X-ray Retrieval**  
- **Multimodal Content Understanding**  

## πŸ“„ Research Paper  
πŸ“œ **arXiv**: [OneEncoder: Progressive Cross-Modal Alignment](https://arxiv.org/abs/2409.11059)  

## πŸ“Œ Resources  
πŸ”— **GitHub Repo**: [OneEncoder](https://github.com/b-faye/OneEncoder)  
πŸš€ **Hugging Face Demo**: [OneEncoder Retriever](https://huggingface.co/spaces/bilalfaye/OneEncoder-retriever)  
πŸ““ **Demo Notebook**: [OneEncoder Demos](https://github.com/b-faye/OneEncoder/tree/main/demo)  
πŸ”Š **OneEncoder for Text, Image**: [HF Model](https://huggingface.co/bilalfaye/OneEncoder-text-image)  
πŸ”Š **OneEncoder for Text, Image & Audio**: [HF Model](https://huggingface.co/bilalfaye/OneEncoder-text-image-audio)  
πŸ”Š **OneEncoder for Text, Image & X-ray**: [HF Model](https://huggingface.co/bilalfaye/OneEncoder-text-image-xray)  


## πŸ“ Authors  
πŸ“Œ **Bilal FAYE**, Hanane AZZAG, Mustapha LEBBAH, Djamel BOUCHAFFRA  

**Note: This model is training with temperature=2.5 and addition as fusion operation**