Salient Information Preserving Adversarial Training Improves Clean and Robust Accuracy
This work addresses the problem of balancing clean and robust accuracy in adversarial training for machine learning models, offering a method that is compatible with human-driven development and provides insights into adversarial risks, though it appears incremental as it builds on existing adversarial training techniques.
The paper tackles the robustness-accuracy trade-off in adversarial training by introducing Salient Information Preserving Adversarial Training (SIP-AT), which uses salient image regions to preserve meaningful features during training, resulting in improved clean accuracy while maintaining high robustness against attacks across multiple datasets and architectures.
In this work we introduce Salient Information Preserving Adversarial Training (SIP-AT), an intuitive method for relieving the robustness-accuracy trade-off incurred by traditional adversarial training. SIP-AT uses salient image regions to guide the adversarial training process in such a way that fragile features deemed meaningful by an annotator remain unperturbed during training, allowing models to learn highly predictive non-robust features without sacrificing overall robustness. This technique is compatible with both human-based and automatically generated salience estimates, allowing SIP-AT to be used as a part of human-driven model development without forcing SIP-AT to be reliant upon additional human data. We perform experiments across multiple datasets and architectures and demonstrate that SIP-AT is able to boost the clean accuracy of models while maintaining a high degree of robustness against attacks at multiple epsilon levels. We complement our central experiments with an observational study measuring the rate at which human subjects successfully identify perturbed images. This study helps build a more intuitive understanding of adversarial attack strength and demonstrates the heightened importance of low-epsilon robustness. Our results demonstrate the efficacy of SIP-AT and provide valuable insight into the risks posed by adversarial samples of various strengths.