Metraipolis › archive › dc_0878a90bf172

DexPolicy: Scheduled Exploration for Trajectory-Guided Dexterous Manipulation — arXiv:2610.00360

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.00360
License (as archived)
CC0-1.0
Archived
2026-10-06T19:16:18.664Z
SHA-256 of the archived text
6ccc56a9b869c09bc0477becabd5a7f70b6119aac52491b28049a333ddae0334
Stable URL of this page
https://metraipolis.com/documento/dc_0878a90bf172

How to cite

DexPolicy: Scheduled Exploration for Trajectory-Guided Dexterous Manipulation — arXiv:2610.00360 — arXiv (en), https://arxiv.org/abs/2610.00360 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0878a90bf172 (Metraipolis archive dc_0878a90bf172, sha256 6ccc56a9b869c09bc0477becabd5a7f70b6119aac52491b28049a333ddae0334)
BibTeX
@misc{dc0878a90bf172,
  title = {DexPolicy: Scheduled Exploration for Trajectory-Guided Dexterous Manipulation — arXiv:2610.00360},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.00360},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0878a90bf172; sha256 6ccc56a9b869c09bc0477becabd5a7f70b6119aac52491b28049a333ddae0334},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.00360. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Reinforcement learning (RL) for dexterous manipulation must discover finger-object contacts and then control the object precisely; the action noise that serves the first goal can interfere with the second. In trajectory-guided settings such as ViViDex, where RL refine hand-object trajectories from human video, our baseline PPO runs end near their initial action noise after 5M steps, motivating explicit control of exploration scale. DexPolicy makes that scale an explicit function of training steps, annealing from broad to narrow exploration while holding loss, architecture, reward, and optimizer settings fixed. We study three policy-optimization settings: PPO, critic-free GRPO continuation, and a flow-parameterized PPO variant (FPO). Across five YCB objects and three training seeds, mean deterministic Target success rises from 49.4% to 68.1% (FPO), 14.1% to 45.4% (GRPO), and 32.0% to 35.7% (PPO). On a RealMan RM75 arm with an Inspire/RH56 hand, 360 trials over three objects raise mean Target success from 25.0% to 85.0% (FPO), 10.0% to 63.3% (GRPO), and 8.3% to 43.3% (PPO), with one trained model per object-method condition. PPO component screening favors noise control over the tested optimizer contraction; the selected PPO schedule yields higher mean Target success than linear decay with the same endpoints on three tested objects. Training return, deterministic Target success, and tolerance to execution noise dissociate; schedules should therefore be judged by terminal task success under the intended execution conditions, per task and policy-optimization setting. Code: this https URL . Website: this https URL .

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-10-02 in cs.CV).
Categorie arXiv: Robotics (cs.RO) ; Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.00360 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-06T17:37:36.921Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-10-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).