Improving Robustness of Vision-Language-Action Models by Restoring Corrupted Visual Inputs

arXiv · 2026-02-01 · Low-Level Learning-Based Action Modelling / Input Modelling / 2D Vision Language Action Models with Robustness / Attack