AI RESEARCH
LatentStealth: Unnoticeable and Efficient Adversarial Attacks on Expressive Human Pose and Shape Estimation
arXiv CS.CV
•
ArXi:2505.12009v2 Announce Type: replace Expressive human pose and shape estimation (EHPS) plays a central role in digital human generation, particularly in live-streaming applications. However, most existing EHPS models focus primarily on minimizing estimation errors, with limited attention on potential security vulnerabilities, such as generating inappropriate content, violent actions, or racially offensive gestures and expressions.