Search papers, labs, and topics across Lattice.
Tongji University, Midea Group
6
0
9
DSWAM bridges the gap between coarse user commands and fine-grained robot actions, outperforming traditional models in real-world task execution.
Injecting 4D geometric priors into World Action Models boosts robotic manipulation performance without adding inference overhead.
Cold start latencies can be slashed by up to 99.3% with CoCoScale's innovative layer-wise scaling approach.
SwiftCache slashes latency by 69% and boosts context length nearly fourfold by enabling cross-model KV cache sharing on GPUs.
PiL-World slashes the gap between simulated and real-world VLA success rates by over 50%, revolutionizing how we evaluate robotic policies.
Radiologists and AI can mutually reinforce each other's diagnostic accuracy by integrating visual cognition traces into a collaborative network for chest X-ray diagnosis.
