Tao Du
About Me
I am a final-year Ph.D. student at MIT CSAIL advised by Wojciech Matusik. My research lies at the intersection of computer graphics, robotics, and machine learning. I develop computational methods to solve challenging design and control problems in complex physical systems. Some of my research topics include building differentiable physics simulation environments for inverse design and control of robots, developing end-to-end pipelines for co-optimizing shape and control of real robots, and exploring continuous Pareto fronts in multi-objective optimization and multi-task learning.
Research Interests
Graphics: computational design and fabrication, physics simulation, shape analysis
Robotics: flying robots, soft robots
Learning: differentiable physics, multi-task learning, sim-to-real transfer
Publications
(* indicates equal contributions)
Functional Optimization of Fluidic Devices with Differentiable Stokes Flow
Tao Du, Kui Wu, Andrew Spielberg, Wojciech Matusik, Bo Zhu, Eftychios Sifakis
SIGGRAPH Asia 2020 (ACM Transactions on Graphics)
Efficient Continuous Pareto Exploration in Multi-Task Learning
Pingchuan Ma*, Tao Du*, Wojciech Matusik
ICML 2020
Learning-in-the-Loop Optimization: End-to-End Control and Co-Design of Soft Robots through Learned Deep Latent Representations
Andrew Spielberg, Allan Zhao, Tao Du, Yuanming Hu, Daniela Rus, Wojciech Matusik
NeurIPS 2019
Learning to Fly: Computational Controller Design for Hybrid UAVs with Reinforcement Learning
Jie Xu, Tao Du, Michael Foshey, Beichen Li, Bo Zhu, Adriana Schulz, Wojciech Matusik
SIGGRAPH 2019 (ACM Transactions on Graphics)
InverseCSG: Automatic Conversion of 3D Models to CSG Trees
Tao Du, Jeevana Priya Inala, Yewen Pu, Andrew Spielberg, Adriana Schulz, Daniela Rus, Armando Solar-Lezama, Wojciech Matusik
SIGGRAPH Asia 2018 (ACM Transactions on Graphics)
Computational Multicopter Design
Tao Du, Adriana Schulz, Bo Zhu, Bernd Bickel, Wojciech Matusik
SIGGRAPH Asia 2016 (ACM Transactions on Graphics)
[Project] [Paper] [Supplement] [Code (design)] [Code (control)] [Hardware]
Convolutional Wasserstein Distances: Efficient Optimal Transportation on Geometric Domains
Justin Solomon, Fernando de Goes, Gabriel Peyré, Marco Cuturi, Adrian Butscher, Andy Nguyen, Tao Du, Leonidas Guibas
SIGGRAPH 2015 (ACM Transactions on Graphics)
Press
Miscellaneous
A Practical Guide to Building a Pentacopter
A step-by-step document that explains how to design, simulate, control, and build a pentacopter for drone hobbyists.