WYLing commited on
Commit
0a360b5
·
verified ·
1 Parent(s): db7f90d

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +89 -3
README.md CHANGED
@@ -1,3 +1,89 @@
1
- ---
2
- license: cc-by-4.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # VisualSimpleQA
2
+
3
+ ## Introduction
4
+ VisualSimpleQA is a multimodal fact-seeking benchmark with two key features. First, it enables streamlined and decoupled evaluation of LVLMs in visual and linguistic modalities. Second, it incorporates well-defined difficulty criteria to guide human annotation and facilitates the extraction of a challenging subset, VisualSimpleQA-hard.
5
+ Experiments on 15 LVLMs show that even state-of-the-art models such as GPT-4o achieve merely 60%+ correctness in multimodal fact-seeking QA on VisualSimpleQA and 30%+ on VisualSimpleQA-hard.
6
+ Furthermore, the decoupled evaluation based on this benchmark across different models highlights substantial opportunities for improvement in both visual and linguistic modules.
7
+
8
+ ## Structure
9
+
10
+ `original_image/`
11
+ This directory contains all image files, where each filename follows the format `original_image_{ID}.png`, matching the unique ID of the corresponding sample in VisualSimpleQA.
12
+
13
+ `cropped_image/`
14
+ This directory contains all cropped rationales from the original images. Each filename follows the format `cropped_image_{ID}.painting`, matching the unique ID of the corresponding sample in VisualSimpleQA.
15
+
16
+ `data.json`
17
+ This JSON file provides detailed information about each sample.
18
+
19
+ `hard_data.json`
20
+ This JSON file provides detailed information about each hard sample in the following format:
21
+
22
+ **Example:**
23
+ ```json
24
+ {
25
+ "id": 369,
26
+ "multimodal_question": "Which institution did the creator of this cartoon duck donate her natural science-related paintings to?",
27
+ "answer": "The Armitt Museum, Gallery, Library",
28
+ "rationale": "Jemima Puddle-Duck",
29
+ "text_only_question": "Which institution did the creator of Jemima Puddle-Duck donate her natural science-related paintings to?",
30
+ "image_source": "https://www.gutenberg.org/files/14814/14814-h/images/15-tb.jpg",
31
+ "evidence": "https://www.armitt.com/beatrix-potter-exhibition/\nhttps://en.wikipedia.org/wiki/Beatrix_Potter",
32
+ "resolution": "400x360",
33
+ "proportion_of_roi": 0.2232,
34
+ "category": "academic and education",
35
+ "text_in_image": "absence",
36
+ "rationale_granularity": "fine-grained"
37
+ }
38
+ ```
39
+
40
+
41
+
42
+ ## Usage
43
+ 1. Ensure the `data.json` file is in the same directory as the script.
44
+ 2. Run the script to randomly select and display a sample.
45
+
46
+ ```python
47
+ import json
48
+ import random
49
+
50
+ with open('./data.json', 'r', encoding='utf-8') as file:
51
+ data = json.load(file)
52
+
53
+ # Randomly select one sample
54
+ random_sample = random.choice(data)
55
+ image_id = random_sample.get('id')
56
+ image_path = f'./original_image/original_image_{image_id}.png'
57
+ multimodal_question = random_sample.get('multimodal_question')
58
+ text_only_question = random_sample.get('text_only_question')
59
+ answer = random_sample.get('answer')
60
+ ```
61
+
62
+ ## Disclaimer
63
+
64
+ This dataset contains images collected from various sources. The authors do NOT claim ownership or copyright over the images. The images may be subject to third-party rights, and users are solely responsible for verifying the legal status of any content before use.
65
+
66
+ - Intended Use: The images are provided for ​non-commercial research purposes only.
67
+
68
+ - Redistribution Prohibition: You may NOT redistribute or modify the images without permission from original rights holders.
69
+
70
+ - Reporting Violations: If you encounter any sample potentially breaching copyright or licensing rules, contact us at [email protected]. Verified violations will be removed promptly.
71
+
72
+ The authors disclaim all liability for copyright infringement or misuse arising from the use of this dataset. Users assume full legal responsibility for their actions.
73
+
74
+ ## License
75
+
76
+ - Text Data: Licensed under [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)
77
+ - Images: Subject to custom terms (see Disclaimer above).
78
+
79
+ ## Citation
80
+
81
+ **BibTeX:**
82
+ ```bibtex
83
+ @article{wang2025visualsimpleqa,
84
+ title={VisualSimpleQA: A Benchmark for Decoupled Evaluation of Large Vision-Language Models in Fact-Seeking Question Answering},
85
+ author={Yanling Wang and Yihan Zhao and Xiaodong Chen and Shasha Guo and Lixin Liu and Haoyang Li and Yong Xiao and Jing Zhang and Qi Li and Ke Xu},
86
+ journal={arXiv preprint arXiv:},
87
+ year={2025}
88
+ }
89
+ ```