Changhe Chen

Northwestern University

Evanston, Illinois

changhec@u.northwestern.edu

I am a CS Ph.D. student at Northwestern University, co-advised by Prof. Manling Li and Prof. Ruohan Zhang. Before Northwestern, I received my M.S. in Robotics from the University of Michigan, where I was advised by Prof. Nima Fazeli and Prof. Xiaonan (Sean) Huang. Prior to Michigan, I received my BASc in Electrical Engineering from the University of Toronto, where I worked with Prof. Chan Carusone on reinforcement learning for analog circuit design. In summer 2025, I collaborated on robotics research and real-world robotic systems with Prof. Heng Yang’s group at Harvard University.

My current research focuses on embodied AI, data-efficient robot learning, and learning from humans, with an emphasis on multimodal vision-language models and instruction-based semantic segmentation for robotic understanding and execution of complex tasks.

/ / changhec [at] u (dot) northwestern (dot) edu

I am open to collaboration. If you have an idea, project, or research direction you would like to discuss, please feel free to contact me by email.

selected publications (*† indicates equal contribution)

RSS

MoTVLA: A Vision-Language-Action Model with Unified Fast-Slow Reasoning

Wenhui Huang^*, Changhe Chen^*, Han Qi, Lv Chen, Yilun Du, and Heng Yang

Robotics: Science and Systems (RSS), 2026

under review

arXiv Bib Paper Website

@article{huang2025motvla,
  title = {MoTVLA: A Vision-Language-Action Model with Unified Fast-Slow Reasoning},
  author = {Huang, Wenhui and Chen, Changhe and Qi, Han and Chen, Lv and Du, Yilun and Yang, Heng},
  year = {2026},
  journal = {Robotics: Science and Systems (RSS)},
  note = {under review},
}

ICRA

Compose by Focus: Scene Graph-based Atomic Skills

Han Qi, Changhe Chen, and Heng Yang

In IEEE International Conference on Robotics and Automation (ICRA), 2026

arXiv Bib Paper Website

@inproceedings{qi2025compose,
  title = {Compose by Focus: Scene Graph-based Atomic Skills},
  author = {Qi, Han and Chen, Changhe and Yang, Heng},
  booktitle = {IEEE International Conference on Robotics and Automation (ICRA)},
  year = {2026},
}

ICRA

ViSA-Flow: Accelerating Robot Skill Learning via Large-Scale Video Semantic Action Flow

Changhe Chen^*, Quantao Yang^*, Xiaohao Xu, Nima Fazeli, and Olov Andersson

In IEEE International Conference on Robotics and Automation (ICRA), 2026

EgoAct workshop, Robotics: Science and Systems (RSS) 2025

arXiv Bib Paper Dataset Website

@inproceedings{chen2025visaflow,
  title = {ViSA-Flow: Accelerating Robot Skill Learning via Large-Scale Video Semantic Action Flow},
  author = {Chen, Changhe and Yang, Quantao and Xu, Xiaohao and Fazeli, Nima and Andersson, Olov},
  booktitle = {IEEE International Conference on Robotics and Automation (ICRA)},
  year = {2026},
  note = {EgoAct workshop, Robotics: Science and Systems (RSS) 2025},
}

RoboSoft

Natural Selection via Foundation Models for Soft Robot Evolution

Changhe Chen^*, Xiaohao Xu^*, Xiangdong Wang^*, Jiaqi Wang, and Xiaonan Huang

IEEE International Conference on Soft Robotics (RoboSoft), 2026

arXiv Bib Paper Code Dataset

@article{chen2025softrobotdesign,
  title = {Natural Selection via Foundation Models for Soft Robot Evolution},
  author = {Chen, Changhe and Xu, Xiaohao and Wang, Xiangdong and Wang, Jiaqi and Huang, Xiaonan},
  year = {2026},
  journal = {IEEE International Conference on Soft Robotics (RoboSoft)},
}

ICRA

CRITERIA: A New Benchmarking Paradigm to Evaluate Trajectory Prediction Approaches

Changhe Chen, Mozhgan Pourkeshavarz, and Amir Rasouli

In IEEE International Conference on Robotics and Automation (ICRA), 2024

Bib Paper Code Dataset

@inproceedings{chen2024criteria,
  title = {CRITERIA: A New Benchmarking Paradigm to Evaluate Trajectory Prediction Approaches},
  author = {Chen, Changhe and Pourkeshavarz, Mozhgan and Rasouli, Amir},
  booktitle = {IEEE International Conference on Robotics and Automation (ICRA)},
  year = {2024},
}

ICCV

Learn TAROT with MENTOR: A Meta-Learned Self-Supervised Approach for Trajectory Prediction

Mozhgan Pourkeshavarz, Changhe Chen, and Amir Rasouli

In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2023

Bib Paper Dataset

@inproceedings{pourkeshavarz2023tarot,
  title = {Learn TAROT with MENTOR: A Meta-Learned Self-Supervised Approach for Trajectory Prediction},
  author = {Pourkeshavarz, Mozhgan and Chen, Changhe and Rasouli, Amir},
  booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
  year = {2023},
}

awards & service

Academic Service

Reviewer, ICRA
Reviewer, IROS

Selective Awards

2020 Fall/Winter Dean's Honours List
2021 Fall/Winter Dean's Honours List
2022 Winter Dean's Honours List
2023 Fall Dean's Honours List
2024 Winter Dean's Honours List
2024 Graduation with honors