Hi, I'm Ying-Chun Lee, and you can also call me Thomas. I'm a master's student in Electrical and Computer Engineering at the University of Washington. I build robot-learning and software infrastructure for real-robot systems, including work around Vision-Language-Action (VLA) models and the tooling that connects policies to hardware: deployment, teleoperation, evaluation, real-time control, and reward modeling.
My robotics experience spans a Robotics Engineer Internship at Chang Chun Group's Information Center, where I built an Isaac Sim digital twin and imitation-learning platform for the ALOHA VX300S arm; Husky Robotics rover perception with OWL-ViT and RealSense depth sensing; and research collaboration with Jiafei Duan and collaborators from Ai2 PRIOR on real-robot policy infrastructure.
Before UW, I conducted research on deep learning-based shoulder ultrasound medical image segmentation with guidance from Prof. Yih-Kuen Jan at the DRES Lab, University of Illinois at Urbana-Champaign (UIUC), and Prof. Chih-Yang Lin. That project became a first-author paper in IEEE Access. In June 2026, I'll join Amazon as a Software Development Engineer Intern.
Focus
- Robot Learning VLA models, reward modeling, simulation, policy learning, and open robot ecosystems.
- Real-Robot Systems Policy deployment, teleoperation, evaluation, and real-time robot infrastructure.
- Chang Chun Group — Information Center Built an Isaac Sim digital twin and imitation-learning tooling for ALOHA VX300S.
- Seattle Source Ranker Distributed data pipeline and React interface over 400K+ GitHub repositories.
Updates
Earlier updates
- Oct 2025 Real-robot policy research with Jiafei Duan and collaborators from Ai2 PRIOR focused on deployment, teleoperation, evaluation, and real-time robot infrastructure.
- Aug 2025 Completed a Robotics Engineer Internship at Chang Chun Group — Information Center, building an imitation-learning platform for the ALOHA VX300S arm.
- Jul 2024 First-author shoulder-ultrasound medical image segmentation paper published in IEEE Access.