Wei Sui

Action ambiguity in navigation is drastically reduced by leveraging embodiment geometry, leading to safer and more effective navigation policies across diverse agents.

TrustVLA can detect and recover from VLA backdoors with minimal clean data, all while preserving normal task performance.

Achieving superior depth estimation accuracy with 88.9% fewer parameters, X-Lens redefines efficiency in heterogeneous camera systems.

Shifting the focus from marginal probabilities to joint trajectory probabilities, dVLA-RL achieves unprecedented success rates in robotic manipulation tasks.

Trajectory preprocessing can boost robotic imitation learning success rates by 25% while cutting down on data size and training expenses.

Humanoid-OmniOcc reveals that a stereo-based dataset can dramatically enhance occupancy prediction accuracy for humanoid robots, outperforming traditional monocular methods.

HoloAgent-0 transforms how robots interpret and act on language instructions, enabling seamless execution of complex tasks in real-world settings.

Key contribution not extracted.