GUANANG SU

GUANANG SU

Student in University of Minnesota

Biography

I am a first year Ph.D. student in Computer Science at University of Minnesota. I work in Robotics:Perception and Manipulation (RPM) Lab with Professor Karthik Desingh. I am passionate about robotics, like vision, manipulation, planning, and perception. My current research topic is robot learning and manipulation with equivariant symmetric.

Before that, I worked as a research assistant in The Helping Hands Lab advised by Professor Robert Platt. Prior to my Ph.D. study, I received my Master’s degree in Robotics from Northeastern University and my Bacholor’s degree in Controls, Robotics, and Autonomy at Virginia Tech and minored in Mathematics.

Willing to be a robotic scientist!

Interests
  • Robotics
  • Imitation Learning
  • Reinforcement Learning
  • Artificial Intelligence
  • Biomedical
Education
  • Ph.D. in Computer Science

    University of Minnesota, Twin City

  • M.S. in Robotics, 2023

    Northeastern University

  • B.S. in Computer Engineering, 2021

    Virginia Polytechnic Institute and State University

Recent Updates

  • 2023.08 Joining RPM Lab and starts my Ph.D. journey!

Publications

(2023). On Robot Grasp Learning Using Equivariant Models. In Autonomous Robots 2023.

PDF Cite Code Video Project Page

On Robot Grasp Learning Using Equivariant Models
(2023). SEIL: Simulation-augmented Equivariant Imitation Learning. In ICRA 2023 (also presented in CoRL 2022 Workshop).

PDF Cite Code Video Project Page

SEIL: Simulation-augmented Equivariant Imitation Learning
(2022). Sample Efficient Grasp Learning Using Equivariant Models. In RSS 2022 (also presented in ICRA 2022 Workshop & RLDM 2022).

PDF Cite Code Video Project Page

Sample Efficient Grasp Learning Using Equivariant Models

Projects

*
Scene Flow Estimation for Autonomous Driving

Scene Flow Estimation for Autonomous Driving

An End-to-end scene flow estimation with PyTorch.

ORB-SLAM3 on iPhone

ORB-SLAM3 on iPhone

ORB SLAM application on mobile device in real-time.

Images Translation with CycleGAN

Images Translation with CycleGAN

Learn the mapping between two unpaired images and apply the image translation in between.

SharkPulse - Shark Image Identification, Dataset Building, Machine Learning

SharkPulse - Shark Image Identification, Dataset Building, Machine Learning

Learn the mapping between two unpaired images and apply the image translation in between.

Skills

🦾
Robotics

Willing be a roboticist!

🎨
Paint / DIY

Enjoy building LEGO / puzzles and painting.

🏸
Sports

Try all, play all.

Blogs

Diffusion Model & Policy

Diffusion Model & Policy

Diffusion policy.