Search papers, labs, and topics across Lattice.
Affiliation:, Nanjing University of Science and Technology
6
0
8
Robots can now reliably act on human gestures by anchoring interaction evidence to specific object instances, achieving over 2x success on implicit intent tasks compared to large VLM baselines.
Human-robot interaction anticipation models exhibit significant directional bias in cue transferability, with pose corrections boosting performance only when transferring from a less to a more visually diverse dataset.
Clinical ML models are only as good as their filtered data, yet current practices obscure crucial scientific decisions that shape task complexity and performance.
Spatial control in co-speech gesture generation can now be fine-tuned in real-time, overcoming the limitations of traditional models that freeze prior motion chunks.
Reward shaping can be the difference between success and failure in optimizing job search queries, with a simple rule-based adjustment yielding a 2.4x performance boost.
HiMPO assigns less-entangled credit to memory updates, significantly boosting long-horizon agent performance while minimizing blame leakage from errors.
