|国家预印本平台
首页|Dynamic Epsilon Scheduling: A Multi-Factor Adaptive Perturbation Budget for Adversarial Training

Dynamic Epsilon Scheduling: A Multi-Factor Adaptive Perturbation Budget for Adversarial Training

Dynamic Epsilon Scheduling: A Multi-Factor Adaptive Perturbation Budget for Adversarial Training

来源:Arxiv_logoArxiv
英文摘要

Adversarial training is among the most effective strategies for defending deep neural networks against adversarial examples. A key limitation of existing adversarial training approaches lies in their reliance on a fixed perturbation budget, which fails to account for instance-specific robustness characteristics. While prior works such as IAAT and MMA introduce instance-level adaptations, they often rely on heuristic or static approximations of data robustness. In this paper, we propose Dynamic Epsilon Scheduling (DES), a novel framework that adaptively adjusts the adversarial perturbation budget per instance and per training iteration. DES integrates three key factors: (1) the distance to the decision boundary approximated via gradient-based proxies, (2) prediction confidence derived from softmax entropy, and (3) model uncertainty estimated via Monte Carlo dropout. By combining these cues into a unified scheduling strategy, DES tailors the perturbation budget dynamically to guide more effective adversarial learning. Experimental results on CIFAR-10 and CIFAR-100 show that our method consistently improves both adversarial robustness and standard accuracy compared to fixed-epsilon baselines and prior adaptive methods. Moreover, we provide theoretical insights into the stability and convergence of our scheduling policy. This work opens a new avenue for instance-aware, data-driven adversarial training methods.

Alan Mitkiy、James Smith、Hana Satou、F Monkey、Hiroshi Tanaka、Emily Johnson

计算技术、计算机技术

Alan Mitkiy,James Smith,Hana Satou,F Monkey,Hiroshi Tanaka,Emily Johnson.Dynamic Epsilon Scheduling: A Multi-Factor Adaptive Perturbation Budget for Adversarial Training[EB/OL].(2025-06-03)[2025-06-14].https://arxiv.org/abs/2506.04263.点此复制

评论