Search papers, labs, and topics across Lattice.
Beijing Humanoid Robot Innovation Center ,, Harbin Institute of Technology , Shenzhen, Carnegie Mellon University
3
0
4
8
A single visual interface, action views, bridges the gap between video generation models and diverse robotic embodiments, enabling unified action modeling and execution.
EvolvingWorld reveals that an open-schema framework can drastically enhance the coherence and depth of character and world interactions in long-horizon literary simulations.
Current visual world models show a dramatic decline in performance when faced with unconventional and impossible physical interactions, highlighting a critical gap in their generalization capabilities.
