Fatemeh Zargarbashi
Robotics, Character animation, Reinforcement learning

PhD Candidate,
ETH Zurich,
Disney research | Studios
Hi! I’m Fatemeh.
I am a Ph.D. candidate in Computer Science jointly at ETH Zurich (Computational Robotics Lab) and Disney research Studios. My research lies in bridging the gap between robotics and character animation by investigating how to develop natural motions on legged robots and how to control physics-based animated characters. I am also interested in developing reinforcement learning algorithms for character control, legged locomotion, and motion generation.
I received my M.Sc. in Mechanical Engineering, Control from Sharif University of Technology, Tehran, Iran. During my master thesis I worked on the development and control of an over-actuated quadcopter for agile maneuvers.
In my free time, I like cycling, swimming and reading poetry. I am also an amateur astronomer and I like to go stargazing whenever I have the chance. Recently, I have started exploring night sky photography
!
news
Jul 05, 2025 | We hosted a booth in the annual ETH RobotX innovation day, with a couple of demos of our research and robots! see post on linkedin |
---|---|
Sep 05, 2024 | I’m thrilled to share that our paper, “RobotKeyframing: Learning Locomotion with High-Level Objectives via Mixture of Dense and Sparse Rewards”, has been accepted at CoRL 2024! 🎉 [arxiv], [youtube] |
latest posts
Jul 08, 2025 | Try DeepMimic in Isaacgym with our assignment |
---|
selected publications
- Learning Steerable Imitation Controllers from Unstructured Animal MotionsarXiv preprint arXiv:2507.00677, 2025
- RobotKeyframing: Learning Locomotion with High-Level Objectives via Mixture of Dense and Sparse Rewards8th Conference on Robot Learning (CoRL), 2024
- MetaLoco: Universal Quadrupedal Locomotion with Meta-Reinforcement Learning and Motion ImitationarXiv preprint arXiv:2407.17502, 2024
- Deep Compliant Control for Legged RobotsIn 2024 IEEE International Conference on Robotics and Automation (ICRA), 2024
- Rl+ model-based control: Using on-demand optimal control to learn versatile legged locomotionIEEE Robotics and Automation Letters, 2023