Yiming Qin

Yiming Qin

PhD student

EPFL

Hello! This is Yiming:

I am a Ph.D. candidate at EPFL, where I am advised by Pascal Frossard. I am currently working on generative models, with a strong passion for graph generation and their potential in drug discovery to advance scientific research. Beyond that, my interests span enhancing control over generative models, graph machine learning, and neural network explainability.

Before starting my PhD, I completed my undergraduate studies in Information Engineering and French at Shanghai Jiao-Tong University (SJTU), within the Paris Élite Institute of Technology (SPEIT). Following this, I was selected for the dual-degree program with École Polytechnique (X), and earned the title of Ingénieur diplômé de l’École polytechnique with a focus on applied mathematics. My studies in France finished with an internship at École Normale Supérieure (ENS), Paris, where I worked on a project titled “Learning Cell Dynamic Features for Unlabeled Video-Microscopies from Synthetic Data” under the supervision Lisa Bedin and Auguste Genovesio. Subsequently, I completed my master’s degree at SJTU under the supervision of Ya Zhang, Jiangchao Yao and Yanfeng Wang.

I’m also passionate about singing! At SJTU, I was part of the a cappella choir in SPEIT, and during my time at École Polytechnique, I played the role of soprano in a Broadway association. Also, I keep painting and I’m also trying to learn guitar.

Download my resumé .

Interests
  • Generative models
  • Graph machine learning
  • Drug discovery
  • Explainability
Education
  • PhD in Computer Science

    École Polytechnique Fédérale de Lausanne (EPFL)

  • ME in Information Engineering, 2023

    Shanghai Jiao-Tong University (SJTU)

  • Ingénieur diplômé de l'École polytechnique (Master level), 2021

    École Polytechnique (X)

  • BE in Information Engineering & French, 2020

    Shanghai Jiao-Tong University (SJTU)

Publications

Quickly discover relevant content by filtering publications.
(2024). DeFoG: Discrete Flow Matching for Graph Generation. arXiv preprint arXiv:2407.08056.

Cite PDF GitHub repository

(2023). Sparse denoising diffusion for large graph generation. arXiv preprint arXiv:2311.02142.

Cite PDF GitHub repository

(2023). Class-balancing diffusion models. CVPR 2023.

Cite PDF GitHub repository

Contact

yiming.qin[at]epfl.ch