Search papers, labs, and topics across Lattice.
Shanghai AI Lab
4
1
7
8
Forget painstakingly collecting real-world data for deformable object manipulation: SIM1 achieves equivalent policy performance with 15x less data by grounding simulations in the physical world.
Reconstructing 3D scenes from images obscured by smoke and extreme darkness is now significantly more achievable, thanks to insights gleaned from the NTIRE 2026 challenge.
Zero-shot visuotactile policies trained in a fast, parallelized simulator can directly control real robots in contact-rich tasks.
VLNVerse tackles the sim-to-real gap in vision-language navigation by providing a unified, large-scale benchmark with realistic physics simulation and full-kinematics embodied agents.
