CV  /  Google Scholar  /  GitHub
Email: zhanpeng at cs.columbia.edu

Zhanpeng He

I am a Ph.D candidate in computer science at Columbia University. My research focuses on robotics, computer vision and reinforcement learning. I am privileged to be co-advised by Professor Matei Ciocarlie and Professor Shuran Song. I am also a member of the Robotic Manipulation and Mobility (ROAM) Lab and the Columbia Artificial Intelligence and Robotics (CAIR) Lab.

Before joining CU, I recieved a master's degree from University of Southern California, where I worked as a research assistant at the Robotic Embedded System Laboratory (RESL) and advised by Professor Gaurav Sukhatme. When I am not working, I enjoy cooking, brewing coffee and playing with my dog Nova. Here is a picture of Nova being corrupted by Gaussian noises.

Research

I'm a full-stack roboticist -- I believe robotics must be solved through progress in many different integral components. While scaling up data collection is important, we also need to design the right hardware (e.g, manipulators and contact sensors) and learning algorithms (e.g., learning from imbalanced sensor data) to enable robust policies on real robots for contact-rich tasks. To this end, my research includes:

My full research statement will be available soon.

News

Publications (* - indicates equal contributions and co-first authors)

Uncertainty Comes Free: Human-in-the-Loop Policies with Diffusion Models

Zhanpeng He*, Yifeng Cao*, Matei Ciocarlie
Under review.
Paper

Meta-World+: An Improved, Standardized, RL Benchmark

Reginald McLean, Evangelos Chatzaroulas, Luc McCutcheon, Frank Röder, Tianhe Yu, Zhanpeng He, KR Zentner, Ryan Julian, JK Terry, Isaac Woungang, Nariman Farsad, Pablo Samuel Castro
Under review. This is an updated version of our previous work Meta-World.
Paper

VibeCheck: Using Active Acoustic Tactile Sensing for Contact-Rich Manipulation

Do-Gon Kim*, Kaidi Zhang*, Hua-Hsuan Liang, Eric T Chang, Zhanpeng He, Ioannis Kymissis, Matei Ciocarlie
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2025
Paper

Task-Based Design and Policy Co-Optimization for Tendon-driven Underactuated Kinematic Chains

Sharfin Islam* Zhanpeng He*, Matei Ciocarlie
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2024
Webpage  /   Paper

MORPH: Design Co-optimization with Reinforcement Learning via a Differentiable Hardware Model Proxy

Zhanpeng He, Matei Ciocarlie
International Conference on Robotics and Automation (ICRA) 2024
Webpage  /   Paper

Decision Making for Human-in-the-loop Robotic Agents via Uncertainty-Aware Reinforcement Learning

Siddharth Singi*, Zhanpeng He*, Alvin Pan, Sandip Patel, Gunnar A. Sigurdsson, Robinson Piramuthu, Shuran Song, Matei Ciocarlie
International Conference on Robotics and Automation (ICRA) 2024
Webpage  /   Paper

Pick2Place: Task-aware 6DoF Grasp Estimation via Object-Centric Perspective Affordance

Zhanpeng He, Nikhil Chavan-Dafle, Jinwook Huh, Shuran Song, Volkan Isler
International Conference on Robotics and Automation (ICRA) 2023
Paper

Discovering Synergies for Robot Manipulation with Multi-Task Reinforcement Learning

Zhanpeng He, and Matei Ciocarlie
International Conference on Robotics and Automation (ICRA) 2022
Webpage  /   Paper   /   Codes

UMPNet: Universal Manipulation Policy Network for Articulated Objects

Zhenjia Xu, Zhanpeng He, and Shuran Song
Robotics and Automation Letters (RA-L)
International Conference on Robotics and Automation (ICRA) 2022
Webpage  /   Paper   /   Codes

Hardware as Policy: Mechanical and ComputationalCo-Optimization using Deep Reinforcement Learning

Tianjian Chen*, Zhanpeng He*, and Matei Ciocarlie
Science Robotics
Conference on Robot Learning (CoRL) 2020
Webpage  /   Paper  /   Science Robotics Article

Learning 3D Dynamic Scene Representations for Robot Manipulation

Zhenjia Xu*, Zhanpeng He*, Jiajun Wu, and Shuran Song
Conference on Robot Learning (CoRL) 2020
Webpage  /   Paper  /   Codes

SQUIRL: Robust and Efficient Learning from Video Demonstration of Long-Horizon Robotic Manipulation Tasks

Bohan Wu, Feng Xu, Zhanpeng He, Abhi Gupta, and Peter K. Allen
International Conference on Intelligent Robots and Systems (IROS) 2020
Webpage  /   Paper  /   Video

Meta-World: A Benchmark and Evaluation for Multi-Task and Meta- Reinforcement Learning

Tianhe Yu*, Deirdre Quillen*, Zhanpeng He*, Ryan C Julian, Karol Hausman, Sergey Levine, and Chelsea Finn
Conference on Robot Learning (CoRL) 2019
Webpage  /   Paper  /   Codes

Simulator Predictive Control: Using Learned Task Representations and MPC for Zero-Shot Generalization and Sequencing

Zhanpeng He*, Ryan C Julian*, Eric Heiden, Hejia Zhang, Stefan Schaal, Joseph Lim, Gaurav S Sukhatme, and Karol Hausman
Neural Information Processing Systems (NIPS) 2018
Paper  /   Codes

Scaling Simulation-to-real Transfer by Learning Composable Robot Skills

Ryan C Julian*, Eric Heiden*, Zhanpeng He, Hejia Zhang, Stefan Schaal, Joseph Lim, Gaurav S Sukhatme, and Karol Hausman
The International Journal of Robotics Research (IJRR)
International Symposium on Experimental Robotics (ISER) 2018
Paper  /   Codes  /   Video  /   Journal version

Mentoring, Service, and Teaching

Currently, at Columbia, I have the luck to work with: Hua-Hsuan Liang (MS in CS), Edward Zhang (UG in CS), and William Wang (UG in CS). Previously, I mentored:

I have been a long-term (>= 4 years) reviewer for several conferences and journals, including: CoRL, ICRA, IROS, RSS, RA-L, TMECH, IJRR, and T-RO. Other than that, I have several teaching experiences:

Softwares

I was a member of rlworkgroup and took part in development of several robot-learning-related open-source projects.

  • Garage: A toolkit for reproducible reinforcement learning research.
  • Dowel: A little logger for machine learning research.
  • Meta-World: A collection of environments for benchmarking meta-learning and multi-task reinforcement learning algorithms.