Search papers, labs, and topics across Lattice.
1
0
3
Robots can now perform intricate assembly tasks and recover from errors in real-time, without any training, by fusing vision-language models with video-based kinematic priors for action planning.
