PS-DLO: Property- and Spatially Aware Deformable Linear Object Manipulation Policy
PS-DLO: Property- and Spatially Aware Deformable Linear Object Manipulation Policy
This paper introduces PS-DLO, a manipulation policy that combines contrastive image-text learning to infer object properties with SAM3-based mask-guided feature extraction for future observations. Evaluated across nine real-world tasks on single-arm and bimanual platforms, the shared multi-task policy achieves an average success rate of 68.9%, surpassing the strongest baseline by 20.0 percentage points. In generalization experiments, PS-DLO outperforms the baseline in six of seven visual-shift conditions for butterfly-knot untying, demonstrating improved robustness to visual variation when combining object-property priors with future spatial supervision.
The authors report a shared multi-task policy trained on nine real-world tasks involving deformable linear objects, achieving an average success rate of 68.9% and outperforming the strongest baseline by 20.0 percentage points; generalization tests show higher success rates than baseline in six of seven visual-shift conditions for butterfly-knot untying.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
Reported success rate
68.9%
Summed sensor-hours
View original evidence
ining a shared multi-task policy for each embodiment. PS-DLO achieves an average success rate of 68.9%, exceeding the strongest baseline by 20.0 percentage points. In generalization experiments, PS-DLO achieves higher success rates than the baseline in six of seven visual-shift conditions on butterfly-knot untying. These results demonstrate that combining object-property priors with future spatial supervision improve
Open source S4tasks evaluated
9
Reported trials
View original evidence
tial observation to categorical descriptions of DLO properties, providing explicit physical and structural priors for action generation. Second, SAM3 extracts mask-guided features from future observations, which are distilled into the policy's action expert to encode task-relevant DLO regions and their subsequent configurations. We evaluate PS-DLO on nine real-world tasks across single-arm and bimanual platforms, tra
Open source S3visual-shift conditions with higher success
6
Reported trials
View original evidence
ining a shared multi-task policy for each embodiment. PS-DLO achieves an average success rate of 68.9%, exceeding the strongest baseline by 20.0 percentage points. In generalization experiments, PS-DLO achieves higher success rates than the baseline in six of seven visual-shift conditions on butterfly-knot untying. These results demonstrate that combining object-property priors with future spatial supervision improve
Open source S4total visual-shift conditions
7
Reported trials
View original evidence
ining a shared multi-task policy for each embodiment. PS-DLO achieves an average success rate of 68.9%, exceeding the strongest baseline by 20.0 percentage points. In generalization experiments, PS-DLO achieves higher success rates than the baseline in six of seven visual-shift conditions on butterfly-knot untying. These results demonstrate that combining object-property priors with future spatial supervision improve
Open source S4
- paper: not_reported
What remains unknown
Not established in the collected evidence: Environment, Control, Data origin.
Reported performance applies to the described task. It does not establish general autonomy or deployment readiness.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
tial observation to categorical descriptions of DLO properties, providing explicit physical and structural priors for action generation. Second, SAM3 extracts mask-guided features from future observations, which are distilled into the policy's action expert to encode task-relevant DLO regions and their subsequent configurations. We evaluate PS-DLO on nine real-world tasks across single-arm and bimanual platforms, tra
Open source S3
ining a shared multi-task policy for each embodiment. PS-DLO achieves an average success rate of 68.9%, exceeding the strongest baseline by 20.0 percentage points. In generalization experiments, PS-DLO achieves higher success rates than the baseline in six of seven visual-shift conditions on butterfly-knot untying. These results demonstrate that combining object-property priors with future spatial supervision improve
Open source S4
Source：arXiv Robotics — research abstracts · arxiv.org