Event
PS-DLO: Property- and Spatial-Aware DLO Manipulation Policy
1 reports · 1 sources
Mengdi Li and other authors propose PS-DLO, a deformable linear object (DLO) manipulation policy that integrates object properties and spatial awareness: it first uses contrastive image-text learning to map initial observations into categorical descriptions of DLO properties, then uses SAM3 to extract mask-guided features from future observations and distills them into an action expert. The authors evaluate on nine real tasks across single-arm and dual-arm platforms, training one shared multi-task policy for each platform. PS-DLO achieves an average success rate of 68.9%, 20.0 percentage points higher than the strongest baseline. Among the seven visual shift conditions in bow-tying, its success rate is higher than the baseline in six.
Report timeline
- PS-DLO: A Property- and Spatially Aware Deformable Linear Object Manipulation Policy
The authors propose PS-DLO, a deformable linear object (DLO) manipulation policy that combines object properties with spatial awareness. Through contrastive image-text learning, it maps initial observations to categorical descriptions of DLO properties, and uses SAM3 to extract mask-guided features from future observations, which are distilled into an action expert. On nine real-world tasks across single-arm and bimanual platforms, PS-DLO achieves an average success rate of 68.9%, 20.0 percentage points higher than the strongest baseline; in six of seven visual-shift conditions for butterfly-knot untying, its success rate exceeds the baseline.
arXiv cs.RO ·