paper-with-me

Papers

Multi-Stage Manipulation with Demonstration-Augmented Reward, Policy, and World Model Learning

2025-03-03 · Adrià López Escoriza, Nicklas Hansen, Stone Tao, Tongzhou Mu, Hao Su

Long-horizon tasks in robotic manipulation present significant challenges in reinforcement learning (RL) due to the difficulty of designing dense reward functions and effectively exploring the expansive state-action space. However, despite a lack of dense rewards, these tasks often have a multi-stage structure, which can be leveraged to decompose the overall objective into manageable subgoals. In this work, we propose DEMO3, a framework that exploits this structure for efficient learning from visual inputs. Specifically, our approach incorporates multi-stage dense reward learning, a bi-phasic training scheme, and world model learning into a carefully designed demonstration-augmented RL framework that strongly mitigates the challenge of exploration in long-horizon tasks. Our evaluations demonstrate that our method improves data-efficiency by an average of 40% and by 70% on particularly difficult tasks compared to state-of-the-art approaches. We validate this across 16 sparse-reward tasks spanning four domains, including challenging humanoid visual control tasks using as few as five demonstrations.

📄 PDF Abstract BibTeX arXiv:2503.01837

Code (1)

adrialopezescoriza/demo3 공식 구현 pytorch

Tasks

Reinforcement Learning (RL)

Similar Papers 제목 키워드 기반

SARM2: Multi-Task Stage Aware Reward Modeling for Self Improving Robotic Manipulation

2026-06-09 · Qianzhong Chen, Hau Zheng, Justin Yu, Suning Huang 외 arxiv

Fine-tuning vision-language-action (VLA) policies for long-horizon manipulation still relies heavily on behavior cloning, which requires costly high-quality demonstrations and keeps policies near the demonstration distri…

Reinforcement Learning

SARM: Stage-Aware Reward Modeling for Long Horizon Robot Manipulation

2025-09-29 · Qianzhong Chen, Justin Yu, Mac Schwager, Pieter Abbeel 외 arxiv

Large-scale robot learning has made progress on complex manipulation tasks, yet long horizon, contact rich problems, especially those involving deformable objects, remain challenging due to inconsistent demonstration qua…

Robot Manipulation

EaDex: A Cross-Embodiment Dexterous Manipulation Framework from Low-Cost Demonstrations

2026-06-02 · Qian Zhao, Xin Tong, Chengdong Wu, Yang Yang 외 arxiv

Dexterous manipulation learning has long been hindered by the high costs of data and training, as pure reinforcement learning typically requires large-scale interactive exploration and imitation learning depends on high-…

Reinforcement Learning

DrS: Learning Reusable Dense Rewards for Multi-Stage Tasks

2024-04-25 · Tongzhou Mu, Minghua Liu, Hao Su

The success of many RL techniques heavily relies on human-engineered dense rewards, which typically demand substantial domain expertise and extensive trial and error. In our work, we propose DrS (Dense reward learning fr…

Robot Manipulation

Stage-Transition Dense Reward Modeling for Reinforcement Learning

2026-06-30 · Yang Yang, Bingjie Chen, Zihan Wang, Yizhe Li 외 arxiv

Reinforcement learning for long-horizon robotic manipulation is often limited by sparse and delayed rewards, while manually designing dense shaping signals is costly and brittle to changes in environments and object conf…

Reinforcement Learning