AugLy: Data Augmentations for Adversarial Robustness
Zoë Papakipos, Joanna Bitton
Abstract
We introduce AugLy, a data augmentation library with a focus on adversarial robustness. AugLy provides a wide array of augmentations for multiple modalities (audio, image, text, & video). These augmentations were inspired by those that real users perform on social media platforms, some of which were not already supported by existing data augmentation libraries. AugLy can be used for any purpose where data augmentations are useful, but it is particularly well-suited for evaluating robustness and systematically generating adversarial attacks. In this paper we present how AugLy works, benchmark it against existing libraries, and use it to evaluate the robustness of various state-of-the-art models to showcase AugLy’s utility. We found that models trained using a wider variety of augmentations were indeed more robust to AugLy augmentations, which validates the hypothesis that training on augmented data improves robustness against adversarial attacks. The AugLy repository can be found at https://github.com/facebookresearch/AugLy.