Self-Labeled Preference Transfer for Steerable Robot Manipulation
Self-Labeled Preference Transfer for Steerable Robot Manipulation
This research proposes Self-Labeled Preference Transfer (SPT), a method that allows robot policies to adapt to new tasks without requiring manual preference annotations. SPT uses a shared vision-language backbone to recognize preferences from annotated source demonstrations and self-labels target demonstrations for fine-tuning. In simulation, SPT achieves 84.8% preference following and 89.2% task success, compared to 59.4% and 88.0% for naive fine-tuning. In real-world experiments, SPT improves preference following from 52.5% to 92.5% relative to naive fine-tuning across two task transfers.
This paper introduces Self-Labeled Preference Transfer (SPT), a method that enables robot policies to adapt to new tasks without manual preference annotations. By leveraging a shared vision-language backbone for preference recognition and self-labeling target demonstrations, SPT achieves improved performance in both simulation and real-world settings compared to naive fine-tuning.
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 accuracy
96.4%
View original evidence
a shared vision-language backbone for preference-conditioned action prediction and preference recognition on annotated source demonstrations. It then freezes the recognizer to preserve its cross-task preference recognition capability and uses it to self-label target demonstrations for policy fine-tuning. Across five types of execution preferences in simulation, the recognizer achieves 96.4% accuracy, and SPT reaches
Open source S4
- 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.
a shared vision-language backbone for preference-conditioned action prediction and preference recognition on annotated source demonstrations. It then freezes the recognizer to preserve its cross-task preference recognition capability and uses it to self-label target demonstrations for policy fine-tuning. Across five types of execution preferences in simulation, the recognizer achieves 96.4% accuracy, and SPT reaches
Open source S4
84.8% preference following and 89.2% task success, compared with 59.4% and 88.0% for naive finetuning. We further extend to multiple preferences, achieving 93.8% preference following compared with 42.5% of baselines. In real-world experiments, SPT increases preference following from 52.5% to 92.5% relative to naive finetuning across two task transfers. These results show that SPT can support preference adaptation ac
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org