Search papers, labs, and topics across Lattice.
1
0
2
Unlabeled human videos can rival traditional datasets in training VLA models, achieving competitive results with minimal annotated data.
