I work on robot learning for legged locomotion and whole-body control.
I earned my MSc in Mechanical Engineering at EPFL, with research experience at EPFL’s BioRob and ETH Zurich’s Robotic Systems Lab.
ANYmal / G1 47 seconds
I study how motion priors can be adapted for legged locomotion and whole-body control. Outside research, I enjoy basketball and hotpot.
MSc in Mechanical Engineering
Project student · With Prof. Marco Hutter
Project student · With Prof. Auke Ijspeert
My research explores how motion priors can be adapted to changing commands, terrain, and physical conditions through parameter optimization, residual policies, and motion generation. Research projects ↗ Paper explainers ↗
RA-L 2026
A diffusion model generates terrain-aware references from human motion, while an RL-trained tracker executes them on a Unitree G1 with onboard perception and computation.
CoRL 2025
A high-level policy combines learned dog locomotion skills with joint-level residual corrections for terrain traversal and local navigation on ANYmal-D.
IROS 2024 · Oral presentation
Online Bayesian optimization adapts a CPG-based controller to changes in velocity commands, ground friction, slope, and payload on a Unitree Go1.
How terrain-aware motion generation and tracking adapt human motion priors for humanoid locomotion.
arXiv 2026 · Under review
An attention-based planner predicts body–environment contacts, which a contact-conditioned controller uses to coordinate whole-body motion.