Pre-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)
- deep-learning
- psychophysics
- human-vision
- object-recognition
132,241 visible repositories in the index
1–3 of 3 · Most stars
Pre-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)
Tidy Plots for Scientific Papers
A simple author handler for scientific writing