Open access
2026
Undermining Trust in Multimodal Computer Vision Models: Realistic Adversarial Patches
It is demonstrated that a once-optimized patch, including patches trained solely on RGB data, retains sufficient attack efficacy to undermine trust in automation systems across the majority of evaluated architectures, even after physically inspired transformations across different multimodal fusion levels.
Max Wiedemann, Luca Eisentraut, Ricardo Buettner
· IEEE Access · 0 citations