Article

Articulated-Body Dynamics Network: Dynamics-Grounded Prior for Robot Learning
Efficient and robust policy learning through novel Articulated-Body Dynamics Network (ABD-NET).
PDF Cite Video
Articulated-Body Dynamics Network: Dynamics-Grounded Prior for Robot Learning
DASH Robot: Minimalistic Design and Optimal Aerial-Terrestrial Locomotion via Contact-Implicit Control
Novel Hybrid Mobile Robot and Locomotion-Mode Implicit Control via MPC.
PDF Cite Video
DASH Robot: Minimalistic Design and Optimal Aerial-Terrestrial Locomotion via Contact-Implicit Control
SPARK: Skeleton-Parameter Aligned Retargeting on Humanoid Robots with Kinodynamic Trajectory Optimization
Human-to-humanoid motion retargeting via kinematics alignment and Kinodynamic trajectory optimization
PDF Cite Video
SPARK: Skeleton-Parameter Aligned Retargeting on Humanoid Robots with Kinodynamic Trajectory Optimization
PRIME: Physically-consistent Robotic Inertial and Motion Estimation for Legged and Humanoid Robots
An optimization based framework to simultaneously estimate robot trajectoties and identify uncertain physical parameters from limited sensory data.
PDF Cite Video
PRIME: Physically-consistent Robotic Inertial and Motion Estimation for Legged and Humanoid Robots
Efficient and Versatile Quadrupedal Skating: Optimal Co-design via Reinforcement Learning and Bayesian Optimization
Optimal Co-design for quadrupedal robotic skating using Bayesian optimization (BO) and Reinforcement Learning (RL)
PDF Cite DOI
Efficient and Versatile Quadrupedal Skating: Optimal Co-design via Reinforcement Learning and Bayesian Optimization
Simultaneous Calibration of Noise Covariance and Kinematics for State Estimation of Legged Robots via Bi-level Optimization
Bi-level optimization framework that jointly calibrates noise covariance and kinematics for state estimation on legged robots.
PDF Cite Code Video DOI
Simultaneous Calibration of Noise Covariance and Kinematics for State Estimation of Legged Robots via Bi-level Optimization