Shuang Li

Postdoctoral Researcher
Stanford University
Departments of EE and CS
Email: lishuang [at] stanford (dot) edu

Google Scholar / Twitter

Research Topics:
Robot Learning
Generative Modeling
Embodied AI

I am a postdoc at Stanford with Shuran Song and Dorsa Sadigh working on Robot Learning. I earned my Ph.D. degree from MIT, under the mentorship of Antonio Torralba. I work closely with Igor Mordatch. My research focuses on Generative Modeling and Robotic Learning.

Recent Talks

[2025/3] Invited talk at AAAI 2025: How Vision and Language Models Are Changing Decision-Making
[2025/2] Course lecture at Stanford CS223A: Perception in Robotics
[2024/11] Course lecture at Stanford EECS227: Pretrained Foundation Models for Robotics
[2024/7] Seminar Talk at Stanford
[2023/12] Invited Talk at NeurIPS 2023 Workshop on Diffusion Models
[2023/8] Invited Talk at Google
[2023/8] Diffusion Models Course at SIGGRAPH
[2023/5] Invited Talk at UC Berkeley
[2023/4] Invited Talk at Nanyang Technological University
[2023/2] Invited Talk at Columbia
[2023/2] Invited Talk at UIUC
[2023/2] Invited Talk at UT Austin
[2022/12] Invited Talk at UC San Diego
[2022/11] Invited Talk at Cornell Tech
[2022/11] Invited Talk at UC Berkeley
[2022/11] Invited Talk at Stanford
[2022/11] Invited Talk at Northeastern

Workshops / Tutorials Organizer

Knowledge in Generative Models at ECCV 2024
Diffusion Models Course at SIGGRAPH 2023
Social Intelligence in Humans and Robots at ICRA 2021, RSS 2022, RSS 2023
MIT Visual Computing Workshop 2021
Wider Face and Person Challenge at ICCV 2019

News

I have been awarded CIFAR AI Chair Award 2024.
I have been selected as one of the EECS Rising Stars 2022.
I have been selected as an Outstanding Reviewer for ICML 2022.
I have been selected as a finalist for 2022 Qualcomm Fellowship. Thanks, Qualcomm!
I have been awarded 2021 Facebook Research Fellowship. Thanks, Facebook!
I have been awarded 2019 Adobe Research Fellowship. Thanks, Adobe!
A summary of our recent works on Compositional Generalization.
A summary of our recent works on Energy-based Models.

Datasets and Environments

VirtualHome multi-agent embodied environment for household activities
V-HICO dataset for video-based human-object interaction detection
CUHK-SYSU dataset for person re-identification
CUHK-PEDES dataset for natural language based person re-identification

Selected Publications [Full List] ( show by date / show by topic)
Research Topics: Robotics Learning / Generative Modeling / Compositional Foundation Models

Selected Honors

CIFAR AI Chair, 2024 [Link]
EECS Rising Stars, 2022 [Link]
ICML Outstanding Reviewer, 2022 [Link]
Qualcomm Fellowship Finalist, 2022 [Link]
Facebook Research Fellowship, 2021 [Link]
Adobe Research Fellowship, 2019 [Link]
MIT Seneff-Zue CS Fellowship, 2018
Outstanding Students Award, CUHK, 2018

Contact

MIT Computer Science and Artificial Intelligence Lab
32 Vassar Street, 32-D394
Cambridge, MA 02139