Grasp-TAGS: Task-Aware Grasp Selection via Critic Distillation for Prehensile Loco-Manipulation
Grasp-TAGS: Task-Aware Grasp Selection via Critic Distillation for Prehensile Loco-Manipulation
This research presents Grasp-TAGS, a framework designed to improve large-object rearrangement by selecting optimal initial grasps for mobile manipulators. By training a unified reinforcement learning policy and distilling its critic into a lightweight network, the system predicts downstream task utility before contact is made. Experiments show an average success rate of 89.84%, outperforming fixed single-grasp policies (79.87%) and reducing orientation errors across various objects.
The paper introduces Grasp-TAGS, a task-aware grasp selection framework for mobile manipulators that uses critic distillation to rank initial grasps based on predicted downstream utility, achieving an average rearrangement success rate of 89.84% compared to 79.87% for fixed single-grasp policies in large-object rearrangement tasks.
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
89.84%
Reported trials
View original evidence
the highest predicted downstream utility for the downstream policy to complete the task, before even making a contact. Experiments demonstrate that Grasp-TAGS improves rearrangement success over both fixed single-grasp policies and random grasp selection, achieving an average success rate of 89.84% compared with 79.87% for the best-performing single-grasp policies, an improvement of 9.97 percentage points. Grasp-TAGS
Open source S5Reported success rate
79.87%
Reported trials
View original evidence
the highest predicted downstream utility for the downstream policy to complete the task, before even making a contact. Experiments demonstrate that Grasp-TAGS improves rearrangement success over both fixed single-grasp policies and random grasp selection, achieving an average success rate of 89.84% compared with 79.87% for the best-performing single-grasp policies, an improvement of 9.97 percentage points. Grasp-TAGS
Open source S5
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: 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.
the highest predicted downstream utility for the downstream policy to complete the task, before even making a contact. Experiments demonstrate that Grasp-TAGS improves rearrangement success over both fixed single-grasp policies and random grasp selection, achieving an average success rate of 89.84% compared with 79.87% for the best-performing single-grasp policies, an improvement of 9.97 percentage points. Grasp-TAGS
Open source S5
also reduces final orientation errors across all evaluated objects. For more qualitative results, refer to the supplementary webpage: https://rpm-lab-umn.github.io/Grasp-TAGS-web-page/
Open source S6
Source：arXiv Robotics — research abstracts · arxiv.org