Robustness Reprogramming for Representation Learning
Zhichao Hou, MohamadAli Torkamani, Hamid Krim, Xiaorui Liu
Abstract
This work tackles an intriguing and fundamental open challenge in representation learning: Given a well-trained deep learning model, can it be reprogrammed to enhance its robustness against adversarial or noisy input perturbations without altering its parameters? To explore this, we revisit the core feature transformation mechanism in representation learning and propose a novel non-linear robust pattern matching technique as a robust alternative. Furthermore, we introduce three model reprogramming paradigms to offer flexible control of robustness under different efficiency requirements. Comprehensive experiments and ablation studies across diverse learning models ranging from basic linear model and MLPs to shallow and modern deep ConvNets demonstrate the effectiveness of our approaches. This work not only opens a promising and orthogonal direction for improving adversarial defenses in deep learning beyond existing methods but also provides new insights into designing more resilient AI systems with robust statistics.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
- Diffusion Models for Adversarial Purification cites Improving Robustness using Generated Data
- Diffusion Models for Adversarial Purification cites Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks
- Improving Robustness using Generated Data cites Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks
- Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks cites Minimally distorted Adversarial Examples with a Fast Adaptive Boundary Attack
- Minimally distorted Adversarial Examples with a Fast Adaptive Boundary Attack cites Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks
Builds on12
- MLP-Mixer: An all-MLP Architecture for VisionIlya O. Tolstikhin, Neil Houlsby, Alexander Kolesnikov, Lucas Beyer et al.NeurIPS 2021 · 3,862 citations
- Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacksFrancesco Croce, Matthias HeinICML 2020 · 2,337 citations
- Diffusion Models for Adversarial PurificationWeili Nie, Brandon Guo, Yujia Huang, Chaowei Xiao et al.ICML 2022 · 663 citations
- Minimally distorted Adversarial Examples with a Fast Adaptive Boundary AttackFrancesco Croce, Matthias HeinICML 2020 · 597 citations
- Improving Robustness using Generated DataSven Gowal, Sylvestre-Alvise Rebuffi, Olivia Wiles, Florian Stimberg et al.NeurIPS 2021 · 384 citations
Related papers
- Transferring Adversarial Robustness Through Robust Representation MatchingPratik Vaishnavi, Kevin Eykholt, Amir RahmatiUSENIX Security 2022
- Word Level Robustness Enhancement: Fight Perturbation with PerturbationPei Huang, Yuting Yang, Fuqi Jia, Minghao Liu et al.AAAI 2022 · 14 citations
- Attack to Explain Deep RepresentationMohammad A. A. K. Jalwana, Naveed Akhtar, Mohammed Bennamoun, Ajmal MianCVPR 2020
- Robust Weight Signatures: Gaining Robustness as Easy as Patching Weights?Ruisi Cai, Zhenyu Zhang, Zhangyang WangICML 2023 · 16 citations
- Measuring Representational Robustness of Neural Networks Through Shared InvariancesVedant Nanda, Till Speicher, Camila Kolling, John P. Dickerson et al.ICML 2022 · 15 citations
