Changan Chen
Co-Founder at Stealth Startup
Palo Alto, California, United States
Summary
🤩
Rockstar🎓
Top SchoolChangan Chen is a research scientist with a decade of experience in robotics and deep reinforcement learning, currently a postdoc at Stanford after earning a PhD from UT Austin. He combines academic rigor with practical ML engineering, contributing to the ICRA’19 CrowdNav project by building CrowdSim environments, adding ORCA navigation policies, and developing reward and testing pipelines. His work spans simulation, algorithm implementation, and empirical evaluation, enabling reproducible experiments that bridge theory and applied navigation systems. Based in China, he brings a global research perspective and a knack for turning complex multi-agent interactions into robust, testable software.
10 years of coding experience
Bachelor’s Degree, Computer Science, Bachelor’s Degree, Computer Science at Zhejiang UniversityDoctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at The University of Texas at AustinPostdoctoral Researcher, Postdoctoral Researcher at Stanford UniversityBachelor’s Degree, Computer Science, Bachelor’s Degree, Computer Science at Simon Fraser University
English, Chinese
Github Skills (3)
machine-learning10
reinforcement-learning10
python10
Programming languages (9)
Github contributions (5)
vita-epfl/CrowdNav
Jul 2018 - Mar 2021
[ICRA19] Crowd-aware Robot Navigation with Attention-based Deep Reinforcement Learning
Role in this project:
ML Engineer
ML EngineerContributions:1 release, 147 commits, 6 PRs in 2 years 8 months
Contributions summary:Changan implemented and modified core components of a crowd-aware robot navigation system based on deep reinforcement learning. They added a new environment class (`CrowdSim`) that defines the simulation environment for agents (pedestrians and a navigator). Further work included adding an `ORCA` policy for robot navigation, and a testing setup, encompassing the creation of testing scenarios and the implementation of reward functions, indicating involvement in the training and evaluation aspects of the navigation models.
robot-navigationreinforcement-learningdeep-reinforcement-learningrobotcollision-avoidance
facebookresearch/sound-spaces
Aug 2020 - Jan 2023
A first-of-its-kind acoustic simulation platform for audio-visual embodied AI research. It supports training and evaluating multiple tasks and applications.
Contributions:3 releases, 82 commits, 1 PR in 2 years 5 months
embodied-aisimulationaudiotrainingacoustic
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.