There are 0 repository under shape-bias topic.
Code to create Stylized-ImageNet, a stylized version of standard ImageNet (ICLR 2019 Oral)
A study on the effect of normalization in predictions by CNN models
Official code for the CVPR 2024 Paper "Can Biases in ImageNet Models Explain Generalization?".
Official repository for the CVPRW2023 paper "An Extended Study of Human-like Behavior under Adversarial Training".
In this work we explore whether MLPs are naturally biased towards using shape or texture as the main feature for image classification decisions.