PAMI: Part Anchored Motion for Text to Human-Object Interaction Generation
PAMI anchors object motion to body parts so generated interactions keep contact and avoid drift.
The paper frames text-to-HOI generation as a coordination problem between full-body motion and object trajectories. Its method has body-part anchors vote on object motion, then refines contact geometry with surface-sensing probes and near-object sensors. On InterAct, the authors report more faithful interactions and 14.5% higher contact recall than the previous state of the art. HF Daily Papers' note
The paper frames text-to-HOI generation as a coordination problem between full-body motion and object trajectories. Its method has body-part anchors vote on object motion, then refines contact geometry with surface-sensing probes and near-object sensors. On InterAct, the authors report more faithful interactions and 14.5% higher contact recall than the previous state of the art. HF Daily Papers' note
score 4