Event
SP-HLM: Sparse Pose-Driven Motion Generation and Whole-Body Control Framework for Humanoid Robots
1 reports · 1 sources
Zhaohui Yang and other researchers propose the SP-HLM framework, which uses five task-critical body poses plus a learned implicit motion prior to represent spatial intent and whole-body coordination, bridging the task policy and whole-body control for humanoid robot loco-manipulation. The paper states that simulation and real humanoid robot experiments show it outperforms existing methods in motion tracking, whole-body coordination, and loco-manipulation success rate. The framework consists of three modules: the SP-VLA vision-language-action model predicts task-relevant sparse poses and finger joint positions via asynchronous action-chunk inference from multimodal observations and language instructions; the SP-MGM motion generation model reconstructs whole-body coordinated motion in real time through two-stage distillation; and the SP-GMT general motion tracker combines explicit sparse targets with the implicit motion prior for whole-body control.
Report timeline
- SP-HLM: Sparse-Pose-Driven Motion Generation and Whole-Body Control Framework for Humanoid Loco-Manipulation
SP-HLM proposes a sparse-pose-driven interface framework that uses five task-critical body poses plus a learned implicit motion prior to represent spatial intent and whole-body coordination in complementary forms. The SP-VLA vision-language-action model first predicts task-relevant sparse poses and finger joint positions from multimodal observations and language instructions via asynchronous action-chunk inference. The SP-MGM motion generation model, trained through two-stage distillation, reconstructs coordinated full-body motion in real time, and the SP-GMT general motion tracker then combines the explicit sparse targets with the implicit motion prior for whole-body control.
arXiv cs.RO ·