Reproduce “Enhancing Adversarial Contrastive Learning via Adversarial Invariant Regularization” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Xilie Xu, Jingfeng Zhang, Feng Liu, Masashi Sugiyama, Mohan Kankanhalli

The authors introduce adversarial invariance regularization (AIR), a penalty added to adversarial contrastive learning (ACL) to make relationships between natural and adversarial image views less dependent on style changes. Added to DynACL, AIR improves standard and adversarial test accuracy on STL-10 under standard linear fine-tuning: 72.30% and 47.66%, respectively, versus 69.59% and 46.49% for DynACL alone.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.