Datasets:

Modalities:
Text
Formats:
csv
Languages:
English
Size:
< 1K
ArXiv:
Libraries:
Datasets
Dask
License:
hemingkx commited on
Commit
b64bc9e
1 Parent(s): a478980

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +54 -1
README.md CHANGED
@@ -9,4 +9,57 @@ tags:
9
  pretty_name: ImageNetVC
10
  size_categories:
11
  - 1K<n<10K
12
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9
  pretty_name: ImageNetVC
10
  size_categories:
11
  - 1K<n<10K
12
+ ---
13
+ # Dataset Card for ImageNetVC
14
+
15
+ ## Dataset Description
16
+
17
+ - **Repository: https://github.com/hemingkx/ImageNetVC**
18
+ - **Paper: https://arxiv.org/abs/2305.15028**
19
+ - **Point of Contact: hemingkx@gmail.com**
20
+
21
+ ### Dataset Summary
22
+
23
+ This is the dataset for our paper "ImageNetVC: Zero-Shot Visual Commonsense Evaluation on 1000 ImageNet Categories".
24
+
25
+ ### Supported Tasks
26
+
27
+ Zero- and Few-shot visual commonsense evaluation.
28
+
29
+ ### Languages
30
+
31
+ English
32
+
33
+ ## Dataset Structure
34
+
35
+ ### Data Instances
36
+
37
+ 4076
38
+
39
+ ### Data Fields
40
+
41
+ Color, shape, material, component, and others.
42
+
43
+ ### Data Splits
44
+
45
+ Dev set for few-shot demonstrations and the main set for zero-shot evaluation.
46
+
47
+ ### Citation Information
48
+
49
+ If you find this dataset useful, please cite our paper:
50
+ ```
51
+ @article{imagenetvc,
52
+ author = {Heming Xia and
53
+ Qingxiu Dong and
54
+ Lei Li and
55
+ Jingjing Xu and
56
+ Ziwei Qin and
57
+ Zhifang Sui},
58
+ title = {ImageNetVC: Zero-Shot Visual Commonsense Evaluation on 1000 ImageNet
59
+ Categories},
60
+ journal = {CoRR},
61
+ volume = {abs/2305.15028},
62
+ year = {2023},
63
+ url = {https://doi.org/10.48550/arXiv.2305.15028},
64
+ }
65
+ ```