prof_pic.jpg

Muyang Li (李沐阳)

EECS, Massachusetts Institute of Technology

I’m a first-year Ph.D. student at MIT, advised by Prof. Song Han. I obtained my master’s degree at Robotics Institute, CMU, advised by Prof. Jun-Yan Zhu and my Bachelor’s degree from Zhiyuan College (ACM Class), Shanghai Jiao Tong University.

My research interest is in the intersection of machine learning, system, and computer graphics. I am currently working on building efficient and hardware-friendly generative models with its applications in computer vision and graphics.

News

Feb 29, 2024 DistriFusion was accepted by CVPR 2024!
Sep 9, 2023 SIGE was accepted by T-PAMI!
Apr 15, 2023 I will join MIT EECS as a Ph.D. student this fall!
Nov 3, 2022 Our new work SIGE has been accepted by NeurIPS 2022! Our code and benchmark are publicly available!
Nov 3, 2022 GAN Compression got 1k stars in Github!

Education

  1. EECS, Massachusetts Institute of Technology
    Ph.D. Student, advised by Prof. Song Han
    Sep. 2023 - Present
  2. Robotics Institute, Carnegie Mellon University
    Master of Science in Robotics, advised by Prof. Jun-Yan Zhu and Prof. Song Han
    Jul. 2021 - May. 2023
  3. Shanghai Jiao Tong University
    Bachelor of Engineering in Computer Science, advised by Prof. Kai Yu and Prof. Yong Yu
    Sep. 2016 - Jul. 2020

Other Experiences

  1. NVIDIA, Shanghai
    Summer Intern, working with Prof. Song Han and Ming-Yu Liu
    Jun 2023 - Aug 2023
  2. OmniML Inc., San Jose
    Summer Intern, working with Prof. Song Han
    May 2022 - Aug 2022
  3. Dawnlight Inc., Shanghai
    Data Scientist, working with Prof. Song Han
    Jul. 2020 - Jul. 2021
  4. MIT HAN Lab, Cambridge
    Research Assistant, working with Prof. Jun-Yan Zhu and Prof. Song Han
    Jul. 2019 - Jan. 2020

Publications

  1. DistriFusion: Distributed Parallel Inference for High-Resolution Diffusion Models
    Muyang Li*, Tianle Cai*, Jiaxin Cao, Qinsheng Zhang, and 6 more authors
    In CVPR 2024
    Website PDF Blog Blog Code
  2. Efficient Spatially Sparse Inference for Conditional GANs and Diffusion Models
    Muyang Li, Ji Lin, Chenlin Meng, Stefano Ermon, and 2 more authors
    In T-PAMI & NeurIPS 2022
    Website PDF Slides Video Code
  3. Lite Pose: Efficient Architecture Design for 2d Human Pose Estimation
    Yihan Wang, Muyang Li, Han Cai, Wei-Ming Chen, and 1 more author
    In CVPR 2022
    PDF Slides Video Code
  4. GAN Compression: Efficient Architectures for Interactive Conditional GANs
    Muyang Li, Ji Lin, Yaoyao Ding, Zhijian Liu, and 2 more authors
    In T-PAMI & CVPR 2020
    Website PDF Slides Video Code

Honors and Awards

  • Honorable Mention, Award for 2019 American Interdisciplinary Contest in Modeling (ICM) Jan, 2019
  • Zhiyuan Honorary Scholarship, (3 times), Award for top 5% students 2016,2017,2018
  • A-Class School-level Scholarship, Award for top 1% students Dec, 2018
  • 1st Runner up’s Coach, Award on ICPC2018 Pathom Regional Nov, 2018
  • Jin Long Yu Fellowship, Award for top 1% students Dec, 2017
  • 1st Runner-up, Award on Singing Competition of Zhiyuan College in SJTU Dec, 2017
  • 3rd Runner-up, Award on ICPC2017 Jakarta Regional Nov, 2017
  • Gold Medal, Award on ICPC2017 Qingdao Regional, Ranked 5th Nov, 2017
  • Gold Medal, Award on CCPC2017 Qingdao Regional, Ranked 10th Oct, 2017

Teaching

  • Coach, SJTU's ICPC team Jun. 2019 - Apr. 2019
  • Teaching Assistant, CS147: Data Structures at SJTU Mar. 2018 - May. 2018

Academic Service

  • Conference Reviewer: ICML, NeurIPS, ICCV, ECCV, CVPR
  • Journal Reviewer: IJCV, TCSVT

Misc.

  • I am enthusiastic about pop singing. I usually publish my singing works at WeSing.