Honours student / researcher

Learning robots to act together.

I work on robot learning, visuomotor imitation, and the systems that help intelligent machines coordinate in the real world.

View publications

Mingyuan Ba photographing beside the water

01 / About

Research starts with a good question.

I am an Honours student building a research profile around learning, perception, and physical interaction.

02 / Research

From demonstrations to capable physical behaviour.

My current work explores how learned policies can help robots coordinate multiple arms and dexterous hands in the real world.

Visuomotor imitation

Learning policies from demonstrations that connect what a robot sees with how it acts.

Multi-arm coordination

Studying how separate manipulators can learn synchronized behaviour across a shared task.

Dexterous manipulation

Using learned hand skills to make contact-rich demonstration collection more reliable and efficient.

03 / Archive

Publications

02 records / 2026

CoRR / arXiv · 2026

NestDex: Nested Policy Learning with Copilot Assisted Teleoperation for Dexterous Manipulation

James Zhao^†, Jinhe Tang^†, Mingyuan Ba, and Weiming Zhi^*

^† Equal contribution. ^* Corresponding author.

A nested policy-learning framework that uses learned hand skills to assist dexterous demonstration collection, then trains an outer visuomotor policy for autonomous deployment.

Read paper

CoRR / arXiv · 2026

Tri-Manual Visuomotor Imitation Learning of Robot Policies

James Zhao^†, Mingyuan Ba^†, and Weiming Zhi^*

^† Equal contribution. ^* Corresponding author.

A tri-manual imitation learning system that lets one operator demonstrate behaviours for three robotic arms. The work introduces Dependency-Aware Tri-Arm Scheduling to retime demonstrations offline before training a synchronized policy.

Read paper

04 / Profile

A work in progress, made public.

This page is a living record of my research interests, publications, and the questions I am working through.

More projects, notes, and a full CV will be added as the work develops.

RoleHonours student

FocusRobot learning

LocationSydney, Australia

05 / Contact

Interested in the same questions?

I am always interested in thoughtful conversations about research, robotics, and new problems worth exploring.

Connect on LinkedIn