Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -12,8 +12,8 @@ dataset_info:
|
|
12 |
dtype: string
|
13 |
- name: filename
|
14 |
dtype: string
|
15 |
-
- name:
|
16 |
-
dtype:
|
17 |
- name: prompt
|
18 |
dtype: string
|
19 |
- name: question
|
@@ -22,8 +22,8 @@ dataset_info:
|
|
22 |
sequence: string
|
23 |
- name: answer
|
24 |
dtype: string
|
25 |
-
- name:
|
26 |
-
dtype:
|
27 |
splits:
|
28 |
- name: test
|
29 |
num_bytes: 1808612125.49
|
@@ -45,12 +45,13 @@ size_categories:
|
|
45 |
---
|
46 |
|
47 |
# π¨ ColorBench
|
|
|
48 |
|
49 |
-
ColorBench is a **multimodal dataset** to comprehensively assess capabilities of VLMs in color understanding, including color perception, reasoning, and robustness, introduced in "ColorBench: Can VLMs See and Understand the Colorful World? A Comprehensive Benchmark for Color Perception, Reasoning, and Robustness".
|
50 |
|
51 |
It provides:
|
52 |
-
- **5,800 image-text questions** covering
|
53 |
-
- **3 categories and 11 tasks** for
|
54 |
|
55 |
## π Instruction
|
56 |
The data/test*.parquet files contain the dataset annotations and images pre-loaded for processing with HF Datasets.
|
|
|
12 |
dtype: string
|
13 |
- name: filename
|
14 |
dtype: string
|
15 |
+
- name: image
|
16 |
+
dtype: image
|
17 |
- name: prompt
|
18 |
dtype: string
|
19 |
- name: question
|
|
|
22 |
sequence: string
|
23 |
- name: answer
|
24 |
dtype: string
|
25 |
+
- name: image_url
|
26 |
+
dtype: string
|
27 |
splits:
|
28 |
- name: test
|
29 |
num_bytes: 1808612125.49
|
|
|
45 |
---
|
46 |
|
47 |
# π¨ ColorBench
|
48 |
+
[**π Paper**](todo) | [**π» GitHub**](https://github.com/tianyi-lab/ColorBench)
|
49 |
|
50 |
+
ColorBench is a **multimodal dataset** to comprehensively assess capabilities of VLMs in color understanding, including color perception, reasoning, and robustness, introduced in ["ColorBench: Can VLMs See and Understand the Colorful World? A Comprehensive Benchmark for Color Perception, Reasoning, and Robustness"](todo).
|
51 |
|
52 |
It provides:
|
53 |
+
- **More than 5,800 image-text questions** covering diverse application scenarios and practical challenges for VLMs evaluation.
|
54 |
+
- **3 categories and 11 tasks** for various color-centric capabilities evaluation including Perception (Color Recognition, Color Extraction and Object Recognition), Reasoning (Color Proportion, Color Comparison, Color Counting, and more.) and Robustness.
|
55 |
|
56 |
## π Instruction
|
57 |
The data/test*.parquet files contain the dataset annotations and images pre-loaded for processing with HF Datasets.
|