Informative Dropout for Robust Representation Learning: A Shape-bias Perspective
This work addresses robustness issues in CNNs for machine learning applications, offering a unified approach to improve performance under various distribution shifts, though it is incremental in building on prior insights about texture-bias.
The paper tackled the problem of convolutional neural networks' reliance on local texture over global shape, which affects robustness, by proposing Informative Dropout (InfoDrop) to reduce texture bias, resulting in enhanced robustness across domain generalization, few-shot classification, image corruption, and adversarial perturbation scenarios.
Convolutional Neural Networks (CNNs) are known to rely more on local texture rather than global shape when making decisions. Recent work also indicates a close relationship between CNN's texture-bias and its robustness against distribution shift, adversarial perturbation, random corruption, etc. In this work, we attempt at improving various kinds of robustness universally by alleviating CNN's texture bias. With inspiration from the human visual system, we propose a light-weight model-agnostic method, namely Informative Dropout (InfoDrop), to improve interpretability and reduce texture bias. Specifically, we discriminate texture from shape based on local self-information in an image, and adopt a Dropout-like algorithm to decorrelate the model output from the local texture. Through extensive experiments, we observe enhanced robustness under various scenarios (domain generalization, few-shot classification, image corruption, and adversarial perturbation). To the best of our knowledge, this work is one of the earliest attempts to improve different kinds of robustness in a unified model, shedding new light on the relationship between shape-bias and robustness, also on new approaches to trustworthy machine learning algorithms. Code is available at https://github.com/bfshi/InfoDrop.