Datasets:
Tasks:
Zero-Shot Classification
Modalities:
Image
Languages:
English
Size:
10K<n<100K
ArXiv:
License:
Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,66 @@
|
|
1 |
-
---
|
2 |
-
license: cc-by-sa-4.0
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: cc-by-sa-4.0
|
3 |
+
task_categories:
|
4 |
+
- zero-shot-classification
|
5 |
+
language:
|
6 |
+
- en
|
7 |
+
pretty_name: ' simco-comco'
|
8 |
+
size_categories:
|
9 |
+
- 10K<n<100K
|
10 |
+
---
|
11 |
+
# **ComCo & SimCo Datasets**
|
12 |
+
|
13 |
+
## **Overview**
|
14 |
+
This repository contains two datasets, **ComCo** and **SimCo**, designed for evaluating multi-object representation in Vision-Language Models (VLMs). These datasets provide controlled environments for analyzing model biases, object recognition, and compositionality in multi-object scenarios.
|
15 |
+
|
16 |
+
- **ComCo**: Composed of real-world objects derived from the COCO dataset.
|
17 |
+
- **SimCo**: Contains simple geometric shapes in structured multi-object settings.
|
18 |
+
|
19 |
+
---
|
20 |
+
|
21 |
+
## **ComCo Dataset**
|
22 |
+
The **ComCo** (Complex COCO Objects) dataset consists of images featuring **2 to 5 objects** from the **COCO dataset**. Each zip file contains different arrangements of objects with variations in:
|
23 |
+
- **Size** (e.g., large vs. small objects)
|
24 |
+
- **Position** (top-left, middle, bottom-right, etc.)
|
25 |
+
|
26 |
+
ComCo is specifically designed to test VLMs on real-world objects, allowing precise control over object placement and ensuring a systematic evaluation of compositional understanding.
|
27 |
+
|
28 |
+
## **SimCo Dataset**
|
29 |
+
The **SimCo** (Simple Compositional Objects) dataset consists of synthetic images featuring **geometric shapes** such as:
|
30 |
+
- **Cubes**
|
31 |
+
- **Spheres**
|
32 |
+
- **Cylinders**
|
33 |
+
- **Triangles**
|
34 |
+
- **Pentagons**
|
35 |
+
|
36 |
+
SimCo is used to **isolate model biases** by removing real-world semantics, enabling controlled evaluation of how VLMs process object interactions purely based on **size, shape, and position**.
|
37 |
+
|
38 |
+
## **Usage**
|
39 |
+
These datasets are useful for:
|
40 |
+
- **Analyzing VLM biases** (e.g., preference for larger objects)
|
41 |
+
- **Compositionality testing** (how models handle multiple objects in images)
|
42 |
+
- **Zero-shot & fine-tuning tasks** (evaluating robustness of vision-language embeddings)
|
43 |
+
|
44 |
+
### **Loading with Hugging Face `datasets` Library**
|
45 |
+
You can load the dataset directly using:
|
46 |
+
```python
|
47 |
+
from datasets import load_dataset
|
48 |
+
|
49 |
+
# Load ComCo dataset
|
50 |
+
comco = load_dataset("clip-oscope/simco-comco", data_dir="ComCo")
|
51 |
+
|
52 |
+
# Load SimCo dataset
|
53 |
+
simco = load_dataset("clip-oscope/simco-comco", data_dir="SimCo")
|
54 |
+
|
55 |
+
```
|
56 |
+
## **Citation**
|
57 |
+
If you use this dataset in your research, please cite:
|
58 |
+
```
|
59 |
+
@inproceedings{abbasi2025clip,
|
60 |
+
title={CLIP Under the Microscope: A Fine-Grained Analysis of Multi-Object Representation},
|
61 |
+
author={Abbasi, Reza and Nazari, Ali and Sefid, Aminreza and Banayeeanzade, Mohammadali and Rohban, Mohammad Hossein and Soleymani Baghshah, Mahdieh},
|
62 |
+
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
|
63 |
+
year={2025}
|
64 |
+
}
|
65 |
+
|
66 |
+
```
|