Yunyi Chen
CS Student & AI Researcher
@ TU Eindhoven
Education
  • TU/e · B.S. CSE
    Sep. 2024 - Present
    GPA: 8.6/10
  • EPFL · Exchange
    Feb. 2026 - Jul. 2026
    GPA: 5.7/6.0
Experience
  • Johns Hopkins University
    Research Intern
    Johns Hopkins University · CCVL Lab
    Jul. 2026 - Present
    Baltimore, MD, USA · Onsite
    Computer Vision · CCVL Lab
  • EPFL
    Semester Project Student
    EPFL · VITA Lab
    Feb. 2026 - Present
    Lausanne, Switzerland · Onsite
    Video Generation · VITA Lab
  • Westlake University
    Research Assistant
    Westlake University · AGI Lab
    Jun. 2025 - Nov. 2025 · 6 mos
    Hangzhou, Zhejiang, China · Onsite
    Diffusion-based Dataset Distillation · AGI Lab
Hiiiiiii! 👋
I'm Yunyi Chen
Computer Science Student & AI Researcher

I am Yunyi Chen (陈云翼), a Computer Science student at Eindhoven University of Technology (TU/e). My research interests span diffusion-based generative models, computer vision, reinforcement learning, and video generation.

I am currently a research intern at the CCVL lab of Johns Hopkins University, working with Prof. Alan Yuille on computer vision. Previously I was on exchange at EPFL (Feb–Jul 2026), joining the VITA lab for a semester project on video generation, supervised by Wuyang Li and Prof. Alexandre Alahi. Before that I was a visiting student at Westlake University (Jun–Nov 2025), where I worked on dataset distillation, resulting in a paper accepted at CVPR 2026 (co-first author) and a first-author submission to NeurIPS 2026.

Outside research I am also a big fan of sports and TV shows, I enjoy playing basketball and table tennis. My favourite TV series are The Big Bang Theory, and Sherlock.

News Scroll for more
Aug 2026
Actively looking for Fall 2027 PhD positions, and remote RA opportunities starting Sep. 2026. Feel free to reach out!
Jun 2026
Started as an on-site Research Intern at Johns Hopkins University, CCVL Lab, working with Prof. Alan Yuille.
Feb 2026
Paper IMS³ accepted at CVPR 2026 as co-first author!
Jan 2026
Started exchange semester at EPFL VITA lab, working on video generation with Prof. Alexandre Alahi and Wuyang Li.
Jan 2026
Submitted MGPO (first author, with Chenru Wang) to NeurIPS 2026.
May 2025
Started as visiting student at Westlake University, Hangzhou, focusing on diffusion-based dataset distillation.
Selected Publications (view all )
IMS³: Breaking Distributional Aggregation in Diffusion-Based Dataset Distillation
CVPR
IMS³: Breaking Distributional Aggregation in Diffusion-Based Dataset Distillation

Chenru Wang*; Yunyi Chen*; Zijun Yang; Joey Tianyi Zhou; Chi Zhang#. (* equal contribution, # corresponding author)

IEEE/CVF Conference on Computer Vision and Pattern Recognition 2026 (Accepted)

Developed Inversion-Matching (IM) to align denoising and inversion trajectories, broadening distributional coverage of distilled samples. Designed Selective Subgroup Sampling (S³) to improve inter-class separability and boundary coverage, achieving state-of-the-art performance among diffusion-based distillation methods.

[arXiv]

IMS³: Breaking Distributional Aggregation in Diffusion-Based Dataset Distillation

Chenru Wang*; Yunyi Chen*; Zijun Yang; Joey Tianyi Zhou; Chi Zhang#. (* equal contribution, # corresponding author)

IEEE/CVF Conference on Computer Vision and Pattern Recognition 2026 (Accepted)

Developed Inversion-Matching (IM) to align denoising and inversion trajectories, broadening distributional coverage of distilled samples. Designed Selective Subgroup Sampling (S³) to improve inter-class separability and boundary coverage, achieving state-of-the-art performance among diffusion-based distillation methods.

[arXiv]

CVPR
MGPO: Manifold-Guided Diffusion Alignment for Task-Aware Dataset Distillation
NeurIPS
MGPO: Manifold-Guided Diffusion Alignment for Task-Aware Dataset Distillation

Yunyi Chen*; Chenru Wang*; Xinyi Ye; Zexin Zheng; Chi Zhang#. (* equal contribution, # corresponding author)

Conference on Neural Information Processing Systems 2026 (Under Review)

Proposed Manifold-Guided Policy Optimization (MGPO), formulating diffusion-based dataset distillation as a multi-objective reinforcement learning problem. Combined pixel-space discriminative rewards with latent-space manifold rewards guided by class-wise MST skeletons, improving downstream classification, detection, and segmentation performance.

[PDF]

MGPO: Manifold-Guided Diffusion Alignment for Task-Aware Dataset Distillation

Yunyi Chen*; Chenru Wang*; Xinyi Ye; Zexin Zheng; Chi Zhang#. (* equal contribution, # corresponding author)

Conference on Neural Information Processing Systems 2026 (Under Review)

Proposed Manifold-Guided Policy Optimization (MGPO), formulating diffusion-based dataset distillation as a multi-objective reinforcement learning problem. Combined pixel-space discriminative rewards with latent-space manifold rewards guided by class-wise MST skeletons, improving downstream classification, detection, and segmentation performance.

[PDF]

NeurIPS
All publications