Awesome
Repository for "What Can Style Transfer and Paintings Do For Model Robustness?" CVPR 2021
Project Website: https://www.cs.cornell.edu/projects/style-painting-robustness/
Please send us an email if you are interested in data or code that has not yet been uploaded!
Datasets
- Materials-60K (OpenSurfaces, MINC, COCO, MIP): <a href=https://www.cs.cornell.edu/projects/style-painting-robustness/datasets/Materials.zip>ZIP [4 GB]</a>
- Out of Distribution Materials (FMD): <a href=https://www.cs.cornell.edu/projects/style-painting-robustness/datasets/FMD_224_all.zip>ZIP [10 MB]</a>
- Objects-1.5K (PACS): <a href=https://www.cs.cornell.edu/projects/style-painting-robustness/datasets/PACS_1499.zip>ZIP [58 MB]</a>
- Out of Distribution Objects (YFCC100M): <a href=https://www.cs.cornell.edu/projects/style-painting-robustness/datasets/flickr_yfcc100m_PACS_categories.zip>ZIP [12 MB]</a>
BibTeX:
@article{lin2021robustness,
title={What Can Style Transfer and Paintings Do For Model Robustness?},
author={Lin, Hubert and van Zuijlen, Mitchell and Pont, Sylvia C and Wijntjes, Maarten WA and Bala, Kavita},
journal={CVPR2021},
year={2021}
}