🤖 Presented at #SPIE2025 in the AI for Security & Defence Applications track our paper "Vision Transformers: the threat of realistic adversarial patches", revealing vulnerabilities in modern AI classification systems.
Vision Transformers show improved robustness over CNNs. However, they remain vulnerable to adversarial patches, with attack success rates ranging from 40% to 99%.
#DFR_ARC #AI #AdversarialAttacks #VisionTransformers #DefenseTech #SPIE #AISecurity
🪖📡 Presented in the Military Cybernetics, Communications & IT, Robotic Systems track at #ICMT2025:
Our paper, “Evaluation Techniques for Modern Military Camouflage” explores how camouflage is assessed in the visible spectrum—a critical area as AI-based detection systems as YOLO continue to evolve
#DFR_ARC investigates how to design & evaluate camouflage that can deceive human & AI observers. Smarter concealment starts with smarter evaluation
🚀 Excited to share that we are developing in the #DFR_ARC project dual-purpose camouflage patterns designed to deceive both human and AI observers. 🕵️♂️
🔍 By investigating vulnerabilities in computer vision systems and examining attacks targeting the visual domain, we create innovative camo designs using computer algorithms. We're partnering with Belgian Defence and #VUB on this PhD research project. 🎓
📢 Stay tuned for updates on our research, publications and the exciting developments ahead! 🌟
Robotics & Autonomous Sytems