Xiang Li

李响 Xiang Li

(He / him)

I'm a Ph.D. candidate in the Dept of CS at Stony Brook University, advised by Prof. Michael S. Ryoo. I received my MS and BS from SJTU under the supervision of Prof. Yue Zhou.

My research interest includes self-supervised visual representation learning, robotics and computer vision.

Email: xiangli8[at]cs[dot]stonybrook[dot]edu

Publication

Crossway Diffusion: Improving Diffusion-based Visuomotor Policy via Self-supervised Learning

New Highlighted

Paper, Code, Video

Xiang Li, Varun Belagali, Jinghuan Shang, and Michael S. Ryoo

Preprint, 2023

Limited Data, Unlimited Potential: A Study on ViTs Augmented by Masked Autoencoders

New

Paper, Code

Srijan Das, Tanmay Jain, Dominick Reilly, Pranav Balaji, Soumyajit Karmakar, Shyam Marjit, Xiang Li, Abhijit Das, and Michael S. Ryoo

WACV, 2024

Peekaboo: Text to Image Diffusion Models are Zero-Shot Segmentors

Paper, Project, Code

Ryan Burgert, Kanchana Ranasinghe, Xiang Li, and Michael S. Ryoo

Preprint, 2023

Related: Diffusion Illusions: Hiding Images in Plain Sight
CVPR Outstanding Demos Award, 2023

Neural Neural Textures Make Sim2Real Consistent

Paper, Project, Code, Video

Ryan Burgert, Jinghuan Shang, Xiang Li, and Michael S. Ryoo

CoRL, 2022

Does Self-supervised Learning Really Improve Reinforcement Learning from Pixels?

Highlighted

Unfortunately, NO (when the same amount of data and augmentation is used).

Paper, Code, Talk

Xiang Li, Jinghuan Shang, Srijan Das, and Michael S. Ryoo

NeurIPS, 2022

StARformer: Transformer with State-Action-Reward Representations for Robot Learning

Paper (TPAMI) (ECCV'22), Code

Jinghuan Shang, Xiang Li, Kumara Kahatapitiya, Yu-Cheol Lee, and Michael S. Ryoo

IEEE TPAMI, Special Issue on Transformer Models in Vision, 2022

Conference version is aceepted by ECCV, 2022

Project

LiDAR based SLAM + GPS + Navigation

ROS C++ Python

This work was done during my internship at Beijing Falcon Self-driving Co., Ltd..

RC Car + Raspi + Camera

Arduino C++ Python

Code

An open-source solution to turn general RC cars into powerful mobile platforms with vision.

CyberDIP: A computer controlled moving finger for touchscreen

C++

Code, Video

With the help of this device and computer vision, we can play mobile games autonomously using the touchscreen on any device.

Visual tracking using a camera on a motorized gimbal

C++

Use a camera on a motorized gimbal and STC algorithm to track moving vehicles. The motorized gimbal is controlled by a PI controller via PELCO-D protocol. GUI is made with MFC.

There are other projects I worked on and enjoyed a lot: playing Android/PC games autonomously using computer vision, GPU cluster monitor built with Grafana and InfluxDB, Single Board Computer (SBC) with various sensors, and so on. They are not technically challenging but I love the fact that they significantly improved the quality of my life :)

Award

CVPR Outstanding Demos Award [Diffusion Illusions], 2023
Outstanding Graduate of Colleges and Universities in Shanghai, China (Top 5%), 2018
National Scholarship for Graduate Students, 2017
Outstanding Graduates of SJTU, 2015

Updated: Nov. 1, 2023 | Back to top