1. Texture Vs ShapePre-trained models, data, code & materials from the paper "ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness" (ICLR 2019 Oral)
2. Stylized ImagenetCode to create Stylized-ImageNet, a stylized version of standard ImageNet (ICLR 2019 Oral)
5. shortcut-perspectiveFigures & code from the paper "Shortcut Learning in Deep Neural Networks" (Nature Machine Intelligence 2020)