Robot Demonstration Generation

Latest papers

13 papers2 papers in the last four weeks, down 60% on the four weeks before. 0.0% of all new papers.

DiscoDemo: Discovering Efficient and Diverse Robot Demonstrations for Imitation Learning
Oct 23, 2026Minho Park, Kinam Kim, Donghu Kim +6Robot Demonstration Generation Robot Imitation Learning
KAIST · Holiday Robotics
Awomo-SimDataEngine: Agentic Simulation-ReadyWorld Generation
Oct 1, 2026Awomo-PhysicalRSI Team, Danjiao Ma, Enhui Ma +21Robot Demonstration Generation Synthetic Data Generation for Robotics
Awomo-PhysicalRSI Team
CAPEX: Efficiently Distilling Foundation Model Behavior into Deployable Robot Policies through Experience-Adaptive Reasoning
Sep 26, 2026Shivam Aarya, Zhang Xi-Jia, Chengyue Huang +6Robot Demonstration Generation Synthetic Data Generation for Robotics
School of Interactive Computing Georgia Institute of Technology Atlanta, GA 30332, USA
DynaForge: Planning-Guided Residual Learning for Dynamic Manipulation Demonstration Generation
Sep 22, 2026Yiyang Jin, Yu Zheng, Xiao He +1Synthetic Data Generation for Robotics Robot Motion Generation
School of Automation and Intelligent Sensing, Shanghai Jiao Tong University, Shanghai 200240, China. · LandSpace Technology Co.,Ltd., China
KnowDemo: Knowledge-Guided Robot Demonstration Generation from Human Videos
Sep 18, 2026Zhiyuan Gao, Yanxiang Zhan, Mohammad Khoshnazar +2Robot Demonstration Generation Synthetic Data Generation for Robotics
University of Bremen · Robotics Institute Germany (RIG)
Demonstration Synthesis from a Single Scan via Gaussian Splatting for Visuomotor Policy Learning
Sep 17, 2026Beichen Wang, Yuen-Hei Yeung, V. R. Sridhar Devarakonda +1Robot Demonstration Generation Synthetic Data Generation for Robotics
George Mason University · New York University
RoboCousin: Build Your Own Simulation Playground for Robust Bimanual Robotic Manipulation
Sep 8, 2026Jingxuan Zhu, Jingyi Li, LiangLiang Chen +3Robotics Simulation Generalization in Robotic Manipulation
E-surfing Digital Life Technology Co., Ltd., China Telecom
PRISM: Personalized Robotic Dataset Generation via Image-based Scene and Motion Synthesis
Jul 6, 2026Dogyu Ko, Haneul Kim, Chanyoung Yeo +3Robotic Data Collection Robot Policy Generalization
Kyung Hee University
JoyAI-Sim: A Simulation-Enabled Interconversion Toolchain for the Embodied Data Pyramid
Jun 15, 2026Peidong Liu, Yongce Liu, Songyan Guo +34Robot Learning Robotic Data Collection
Joy Future Academy, JD Group
RoboDream: Compositional World Models for Scalable Robot Data Synthesis
Jun 1, 2026Junjie Ye, Rong Xue, Basile Van Hoorick +6Robot Skill Learning Robot Demonstration Generation
1USC Physical Superintelligence (PSI) Lab · 2Toyota Research Institute
RDGen: Demonstration Generation for High-Quality Robot Learning via Reinforcement Learning
May 29, 2026Zijian Zhu, Menglin Zou, Zhuang Li +2RL for Robotics Robot Skill Learning
AffordGen: Generating Diverse Demonstrations for Generalizable Object Manipulation with Afford Correspondence
Apr 12, 2026Jiawei Zhang, Kaizhe Hu, Yingqian Huang +3Affordance-Based Robotic Manipulation Robotic Manipulation
Shanghai Qi Zhi Institute · Tsinghua University · Fudan University +1
Scene2Demo: Self-Evolving Embodied Data Generation via Object-Action Graph
Feb 12, 2026Xiang Liu, Sen Cui, Guocai Yao +4Robot Demonstration Generation Synthetic Data Generation for Robotics

Oct 23, 2026J/K move · Enter open · S save

DiscoDemo: Discovering Efficient and Diverse Robot Demonstrations for Imitation Learning

Minho Park, Kinam Kim, Donghu Kim, Byungkun Lee+5

KAIST · Holiday Robotics

Robot Demonstration Generation Robot Imitation Learning Synthetic Data Generation for Robotics Robot Manipulation Datasets RL for Robotics

Open paper AskPDF arXiv

Behavior cloning for robotic manipulation is highly sensitive to the coverage of its training data, yet collecting diverse successful demonstrations remains costly. We study reinforcement learning (RL) not as a deployable controller, but as a data generator for downstream visuomotor imitation learning. Our key premise is that multiple successful solution modes cover a broader range of successful trajectories than repeatedly collecting a single narrow solution. We introduce DiscoDemo, which trains a privileged latent-conditioned RL generator with a parallel reverse curriculum and a success-gated diversity objective to discover diverse solutions while preserving task success. The trained generator produces multiple successful closed-loop behaviors, which are safety-filtered and exported as a static dataset for downstream imitation learning. Across four manipulation tasks, DiscoDemo produces broader successful-state coverage and improves downstream visuomotor policies in both simulation and the real world. Averaged across tasks, downstream success increases to 62% from 40% for the strongest baseline in simulation, and to 78% from 39% on the real robot, respectively. These results suggest that successful solution diversity improves imitation learning for a pretrained flow-matching VLA and remains complementary to corrective data collection. Project page: https://davian-robotics.github.io/DiscoDemo