Search papers, labs, and topics across Lattice.
1
0
3
7
Pre-trained video diffusion models can be deterministically adapted into state-of-the-art zero-shot depth estimators, sidestepping the need for massive labeled datasets.
