5 papers
HORIZON: Recoverability-Governed Curriculum for Physical-Domain Scaling
Chenhao Bai, Liqin Lu, Kaijun Wang +5
Scaling robust robot policies requires more than broader randomization, because physical-domain experience must remain organized and learnable throughout training. We study when a…
Translating Flow to Policy via Hindsight Online Imitation
Yitian Zheng, Zhangchen Ye, Weijun Dong +5
Recent advances in hierarchical robot systems leverage a high-level planner to propose task plans and a low-level policy to generate robot actions. This design allows training the…
TimeRewarder: Learning Dense Reward from Passive Videos via Frame-wise Temporal Distance
Yuyang Liu, Chuan Wen, Yihang Hu +2
Designing dense rewards is crucial for reinforcement learning (RL), yet in robotics it often demands extensive manual effort and lacks scalability. One promising solution is to vie…
Prior Reinforce: Goal-Conditioned Dynamic Manipulation with Limited Trials
Yihang Hu, Pingyue Sheng, Yuyang Liu +2
Embodied robots have achieved strong performance in many real-world manipulation tasks, yet agile dynamic manipulation remains challenging due to high sensitivity to motion paramet…
Imitation Learning from Observation with Automatic Discount Scheduling
Yuyang Liu, Weijun Dong, Yingdong Hu +4
Humans often acquire new skills through observation and imitation. For robotic agents, learning from the plethora of unlabeled video demonstration data available on the Internet ne…