Search papers, labs, and topics across Lattice.
Affiliation:
5
1
8
4
This work introduces physical rank consistency (PRC) to measure how well tokenization preserves local physical distance rankings after reconstruction, and presents ActionPiece, which preserves physical action relationships through joint supervision of representation learning and quantization.
Qualitative examples show the PhysBrain 1.5 model's ability to produce end-effector trajectories and predict future scenes through spatially aligned RGB, depth, and robot-mask outputs.
SIEVE reveals that leveraging reusable structures in demonstration data can lead to more efficient and effective imitation learning, outperforming full-data training with significantly less input.
Human-as-Humanoid achieves a staggering 4.8–7.2x increase in demonstration throughput, transforming how humanoid robots learn from human actions.
Achieve >97.5% of full-data VIT performance with only 16% of the data using ScalSelect, a surprisingly effective and scalable training-free data selection method.
