defense arXiv Mar 30, 2026 · 9d ago
Ziad Sharawy, Mohammad Nakshbandiand, Sorin Mihai Grigorescu · Transylvania University of Brașov
Statistical detection framework for adversarial attacks on semantic segmentation models in robotic perception systems
Input Manipulation Attack vision
Deep Neural Networks (DNNs) achieve strong performance in semantic segmentation for robotic perception but remain vulnerable to adversarial attacks, threatening safety-critical applications. While robustness has been studied for image classification, semantic segmentation in robotic contexts requires specialized architectures and detection strategies.
cnn Transylvania University of Brașov
ML Security Papers