Search papers, labs, and topics across Lattice.
4
0
3
2
TAMP-Nav aligns embodied navigation with VLMs' 2D capabilities, achieving a 66.2% success rate while drastically improving efficiency.
Asynchronous tracking for humanoid robots can significantly enhance performance and safety in dynamic environments, outperforming traditional methods even at low update rates.
APT achieves substantial improvements in instruction generalization for VLA models by effectively decoupling language and action learning, addressing a critical data imbalance issue.
Achieve robust 3D shape recovery across viewpoints and unseen object categories by explicitly modeling and correlating template geometry with target observations.
