Search papers, labs, and topics across Lattice.
1
0
2
Current video generation models fail to maintain embodiment consistency and functional interaction in human-to-robot manipulation, revealing significant gaps in their transfer capabilities.
