Search papers, labs, and topics across Lattice.
4
0
3
0
Humanoid robots can now learn complex loco-manipulation by transferring visual-motion priors from over 10,000 hours of human data, achieving top performance on simulated and real-world tasks.
PLAT bridges dense motion tracking and sparse goal-conditioned control by exploiting dense goal sequences as privileged supervision during training while requiring only sparse timed keyframe commands at deployment.
The challenge revealed that innovative techniques can significantly enhance visibility for autonomous vehicles in rainy conditions, a key hurdle for safe navigation.
GRNEdit achieves superior video editing performance with significantly fewer parameters, challenging the notion that bigger models are always better.
