Adversarial Vulnerabilities of Learned Telesurgery Policies
The episode discusses a paper on adversarial vulnerabilities in learned telesurgery policies, focusing on disruptive and steering attacks that can cause harm. Hosts discuss how these threats test system resilience across different architectures like ACT and Diffusion Policy. Key takeaways include the need for real-time detection mechanisms and developing adaptive defenses like Temporal Photometric Attack (TPA) to ensure surgical robot safety.