Search papers, labs, and topics across Lattice.
3
0
4
0
Visual RL agents can recover near-perfect performance even under severe, dynamically changing visual corruptions by learning to disentangle task-relevant foreground from perturbation artifacts.
By cleverly embedding the face itself as a watermark within the background, RecoverMark offers a surprisingly robust defense against face manipulation, even when attackers try to remove the watermark.
A novel adversarial attack, MVIG, can reduce the success rate of state-of-the-art defenses in collaborative perception systems by up to 62%, revealing critical security gaps.
