Ziru Wang

MonoLift: Learning 3D Manipulation Policies from Monocular RGB via Distillation

Pre-Training Robo-Centric World Models For Efficient Visual Control

ReDebias: Exploring Residual energy based Debias learning

DynaMind: Reasoning over Abstract Video Dynamics for Embodied Decision-Making

Imagine Within Practice: Conservative Rollout Length Adaptation for Model-Based Reinforcement Learning