Chenyang Xu

Achieving a 22.3% word error rate with just 240 ms latency, LipsFlow redefines the capabilities of Visual Speech Recognition in challenging multi-speaker environments.

Neuromorphic event cameras can unlock a new level of expressiveness in speech generation, capturing subtle emotional nuances missed by traditional RGB-based methods.

Unlock generalist robots by learning manipulation skills directly from the abundance of human activity videos, bypassing the robot data bottleneck.