Datasets:

Modalities:
Image
Text
Formats:
parquet
ArXiv:
Libraries:
Datasets
Dask
License:
Fbordes commited on
Commit
c06d287
1 Parent(s): 00bb5e2

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +22 -0
README.md CHANGED
@@ -30,3 +30,25 @@ configs:
30
  - split: train
31
  path: data/train-*
32
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
30
  - split: train
31
  path: data/train-*
32
  ---
33
+
34
+ ## PUG: SPAR
35
+ PUG: SPAR (Scene, Position, Attribute, Relation) contains 43,560 test samples, with image-caption pairs that evaluate VLMs scene and object recognition, as well as inter-object and object-attribute relationships respectively. We utilize scenes containing up to two objects in 4 unique spatial relationships and 4 different texture variations.
36
+
37
+ ## LICENSE
38
+ The datasets are distributed under the CC-BY-NC, with the addenda that they should not be used to train Generative AI models.
39
+
40
+ ## Citing PUG
41
+ If you use one of the PUG datasets, please cite:
42
+ ```
43
+ @misc{bordes2023pug,
44
+ title={PUG: Photorealistic and Semantically Controllable Synthetic Data for Representation Learning},
45
+ author={Florian Bordes and Shashank Shekhar and Mark Ibrahim and Diane Bouchacourt and Pascal Vincent and Ari S. Morcos},
46
+ year={2023},
47
+ eprint={2308.03977},
48
+ archivePrefix={arXiv},
49
+ primaryClass={cs.CV}
50
+ }
51
+ ```
52
+
53
+ ## To learn more about the PUG datasets:
54
+ Please visit the [website](https://pug.metademolab.com/) and the [github](https://github.com/facebookresearch/PUG)