Erosion Attack for Adversarial Training to Enhance Semantic Segmentation Robustness
This addresses robustness issues in semantic segmentation models for computer vision applications, representing an incremental improvement over existing adversarial training approaches.
The paper tackles the vulnerability of semantic segmentation models to adversarial attacks by proposing EroSeg-AT, a framework that uses EroSeg to generate adversarial examples by disrupting semantic consistency through pixel-level perturbation propagation, resulting in significantly improved attack effectiveness and model robustness compared to existing methods.
Existing segmentation models exhibit significant vulnerability to adversarial attacks.To improve robustness, adversarial training incorporates adversarial examples into model training. However, existing attack methods consider only global semantic information and ignore contextual semantic relationships within the samples, limiting the effectiveness of adversarial training. To address this issue, we propose EroSeg-AT, a vulnerability-aware adversarial training framework that leverages EroSeg to generate adversarial examples. EroSeg first selects sensitive pixels based on pixel-level confidence and then progressively propagates perturbations to higher-confidence pixels, effectively disrupting the semantic consistency of the samples. Experimental results show that, compared to existing methods, our approach significantly improves attack effectiveness and enhances model robustness under adversarial training.