TransPatch: Learning a Universal Adversarial Patch for ViT–CNN Cross-Architecture Transfer in Semantic Segmentation (Student Abstract)

Robust semantic segmentation is crucial for safe autonomous driving, yet deployed models remain vulnerable to black-box adversarial attacks when target weights are unknown. Most existing approaches either craft image-wide perturbations or optimize patches for a single architecture, which limits their practicality and transferability. We introduce TransPatch, a training framework for learning a universal adversarial patch that generalizes across images and architectures without requiring access to target model parameters.