Marina Y. Aoyama

I am a fourth-year Ph.D. student in Robotics at the University of Edinburgh. I am fortunate to be advised by Prof. Sethu Vijayakumar and Prof. Subramanian Ramamoorthy. My research interests lie in robotics, focusing on contact-rich manipulation, imitation learning, reinforcement learning, sim-to-real transfer, and force and tactile sensing.

Specifically, I develop algorithms that enable robots to learn physically interactive tasks from limited real-world data by leveraging simulation and self-supervised learning. My goal is to develop robots capable of generalizing manipulation skills across diverse object properties and contact scenarios through active interaction and the use of force and tactile feedback.

I am currently an intern at Google DeepMind, working with Todor Davchev. Previously, I worked with Tetsuya Narita as an intern at Sony Group Corporation.

Ph.D. Candidate in Robotics

Statistical Learning and Motor Control Group SLMC

The University of Edinburgh

Email: m.aoyama[at]ed.ac.uk

[Google Scholar][github]

Research Interests

Efficient Learning for Contact-Rich Manipulation & Physical Human-Robot Interaction

Designing data-efficient sim-to-real transfer methods for manipulating objects with diverse physical properties.
Developing transferable object and skill representations for cross-task and cross-object generalisation.
Optimising demonstration and interaction data collection using explainable models and uncertainty quantification.
Integrating model-based and data-driven strategies for sample-efficient manipulation learning.
Implementing and validating proposed frameworks on robotic hardware.

Keywords: Contact-rich manipulation, Learning from Demonstration, Reinforcement Learning, Force and tactile sensing, Representation learning, Transfer learning, Assistive robotics.

Page updated

Google Sites

Report abuse