Search papers, labs, and topics across Lattice.
1
0
3
Monolithic 3D scenes with hundreds of heavily occluded objects can be cleanly parsed into individual editable meshes without ever training on multi-object data.