Search papers, labs, and topics across Lattice.
3
0
3
15
SDM reveals that pretrained image models can disentangle camera and object motion in videos, outperforming traditional methods with minimal supervision.
Video generation models could be the key to unlocking general-purpose vision intelligence, outperforming specialized models with far less training data.
Ditch slow, 2D motion proxies: GMOS directly segments moving objects from RGB video in 3D space and time, achieving state-of-the-art speed and accuracy.
