Search papers, labs, and topics across Lattice.
4
0
5
Disentangling ego-motion from environmental dynamics allows FR3D to achieve unprecedented geometric consistency in future 3D reconstructions.
A compact scene graph representation allows for efficient visual localization, slashing storage needs while preserving accuracy in complex environments.
Label-free consistency training can elevate spatial reasoning in LRMs to levels comparable with traditional supervised methods, revolutionizing how we approach model training.
DynaTok achieves superior 4D reconstruction from incomplete point clouds, improving both quality and temporal coherence without relying on images or explicit correspondences.
