Effective Targeted Attacks for Adversarial Self-Supervised Learning

Sung Hwang, Minseon Kim, Hyeonjeong Ha, Sooel Son

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 1 functions out of this paper's own repositories and ran 0 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

Repositories linked to this paper

• Kim-Minseon/TARO

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Recently, unsupervised adversarial training (AT) has been highlighted as a means of achieving robustness in models without any label information. Previous studies in unsupervised AT have mostly focused on implementing self-supervised learning (SSL) frameworks, which maximize the instance-wise classification loss to generate adversarial examples. However, we observe that simply maximizing the self-supervised training loss with an untargeted adversarial attack often results in generating ineffective adversaries that may not help improve the robustness of the trained model, especially for non-contrastive SSL frameworks without negative examples. To tackle this problem, we propose a novel positive mining for targeted adversarial attack to generate effective adversaries for adversarial SSL frameworks. Specifically, we introduce an algorithm that selects the most confusing yet similar target example for a given instance based on entropy and similarity, and subsequently perturbs the given instance towards the selected target. Our method demonstrates significant enhancements in robustness when applied to noncontrastive SSL frameworks, and less but consistent robustness improvements with contrastive SSL frameworks, on the benchmark datasets.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2210.10482")
get_code_for_paper("2210.10482")
have("2210.10482")

Connect an agent — have() is free.