ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 2303.10225
21
0

Robust Mode Connectivity-Oriented Adversarial Defense: Enhancing Neural Network Robustness Against Diversified ℓp\ell_pℓp​ Attacks

17 March 2023
Ren Wang
Yuxuan Li
Sijia Liu
    AAML
ArXivPDFHTML
Abstract

Adversarial robustness is a key concept in measuring the ability of neural networks to defend against adversarial attacks during the inference phase. Recent studies have shown that despite the success of improving adversarial robustness against a single type of attack using robust training techniques, models are still vulnerable to diversified ℓp\ell_pℓp​ attacks. To achieve diversified ℓp\ell_pℓp​ robustness, we propose a novel robust mode connectivity (RMC)-oriented adversarial defense that contains two population-based learning phases. The first phase, RMC, is able to search the model parameter space between two pre-trained models and find a path containing points with high robustness against diversified ℓp\ell_pℓp​ attacks. In light of the effectiveness of RMC, we develop a second phase, RMC-based optimization, with RMC serving as the basic unit for further enhancement of neural network diversified ℓp\ell_pℓp​ robustness. To increase computational efficiency, we incorporate learning with a self-robust mode connectivity (SRMC) module that enables the fast proliferation of the population used for endpoints of RMC. Furthermore, we draw parallels between SRMC and the human immune system. Experimental results on various datasets and model architectures demonstrate that the proposed defense methods can achieve high diversified ℓp\ell_pℓp​ robustness against ℓ∞\ell_\inftyℓ∞​, ℓ2\ell_2ℓ2​, ℓ1\ell_1ℓ1​, and hybrid attacks. Codes are available at \url{https://github.com/wangren09/MCGR}.

View on arXiv
Comments on this paper