Zeji Yi

I am a second-year Ph.D. student in Electrical and Computer Engineering at Carnegie Mellon University, where I am affiliated with the Robotics Institute. I am fortunate to be advised by Prof. Guannan Qu and Guanya Shi(LeCAR Lab). Previously, I recieved my master’s degree at Tsinghua University, advised by Prof.Yanan Sui(LNS Group). I also hold a bachelor’s degree with honors from the Tsien Excellence in Engineering Program at Tsinghua University.

Email  /  CV  /  Scholar  /  Twitter  /  Github

Research

I am broadly interested in the intersection of learning, control, and optimization. My research focuses on developing theoretical tools to deepen our understanding of the connections between learning and control, with the aim of inspiring new algorithms that are more applicable to real-world robotic tasks.

DIAL-MPC is the first training-free method achieving real-time whole-body torque control using full-order dynamics.

We introduce a safe and efficient high-dimensional Bayesian optimization approach for complex control tasks including musculoskeletal systems.

MBD is a diffusion-based trajectory optimization method that directly computes the score function using models without any external data.

We quantify the convergence rate of sampling-based MPC, and design a practical and effective algorithm CoVO-MPC with optimal rate.

We utilize MCMC and Langevin Dynamics to efficiently propose sampling points for Bayesian Optimization in High-Dimensional space

Covariance control for stochastic system with stochastic differential dynamic programming.