Muntasir Wahed, Afrina Tabassum, Ismini Lourentzou

Abstract

Contrastive learning has gained popularity as an effective self-supervised representation learning technique. Several research directions improve traditional contrastive approaches, e.g., prototypical contrastive methods better capture the semantic similarity among instances and reduce the computational burden by considering cluster prototypes or cluster assignments, while adversarial instance-wise contrastive methods improve robustness against a variety of attacks. To the best of our knowledge, no prior work jointly considers robustness, cluster-wise semantic similarity and computational efficiency. In this work, we propose SwARo, an adversarial contrastive framework that incorporates cluster assignment permutations to generate representative adversarial samples. We evaluate SwARo on multiple benchmark datasets and against various white-box and black-box attacks, obtaining consistent improvements over state-of-the-art baselines.

People

Ismini Lourentzou


Afrina Tabassum


Muntasir Wahed


Publication Details

Date of publication:
April 21, 2022
Journal:
Cornell University
Publication note:

Muntasir Wahed, Afrina Tabassum, Ismini Lourentzou: Adversarial Contrastive Learning by Permuting Cluster Assignments. CoRR abs/2204.10314 (2022)