黎子韵

我是黎子韵(Ziyun Li),现为瑞典皇家理工学院(KTH)生成式AI方向的博士后研究员,驻地斯德哥尔摩。

我的研究方向涵盖生成式AI流匹配(Flow Matching)、扩散模型(Diffusion Models)、公平性(Fairness)、AI安全(AI Safety)以及AI与社会

最新动态

2026年5月 我们的论文《A Kinetic-Energy Perspective of Flow Matching》被 ICML 2026 接收为亮点论文(Spotlight,Top 2.2%)!🎉
2026年5月 上线 YunFlow, AI Agent,覆盖 AI 前沿与技术、全球市场、创投与预测市场,每日结构化简报(可邮件订阅)。代码 🎉
2025年2月 开始在瑞典皇家理工学院(KTH)担任生成式AI方向博士后研究员,由Wallenberg AI、自主系统与软件项目(WASP)资助!🎉
2024年11月 成功通过哈索·普拉特纳研究院(HPI) AI/ML 方向博士论文答辩!期待开启科研新旅程。

代表论文

2026

  1. ICML Spotlight
    A Kinetic-Energy Perspective of Flow Matching
    Ziyun Li, Huancheng Hu, Soon Hoe Lim, and 6 more authors
    In International Conference on Machine Learning (ICML), 2026
    Spotlight Presentation [Top 2.2%]
  2. AsiaCCS
    LotBoNC: Novel Botnet Traffic Classification under Long-tailed Distributions
    Huancheng Hu, Ziyun Li, and Christian Doerr
    In ACM Asia Conference on Computer and Communications Security (AsiaCCS), 2026

2025

  1. EurIPS
    EnfoPath: Energy-Informed Analysis of Generative Trajectories in Flow Matching
    Ziyun Li, Ben Dai, Huancheng Hu, and 2 more authors
    In EurIPS Workshop, 2025
  2. COPA
    FlowGuard: Guarding Flow Matching via Conformal Sampling
    Ziyun Li, and Henrik Boström
    In Conformal and Probabilistic Prediction With Applications (COPA), 2025
  3. OJCOMS
    BoNC: Discovering and Classifying Novel Encrypted Botnet Traffic
    Huancheng Hu, Ziyun Li, and Christian Doerr
    Open Journal of Communication Societies (OJCOMS), 2025
  4. ICCV
    Q-Norm: Robust Representation Learning via Quality-Adaptive Normalization
    Ying Zhou, Lanning Zhang, Fei Gao, Ziyun Li, Maoying Qiao, Jinlan Xu, and Nannan Wang
    In International Conference on Computer Vision (ICCV), 2025
  5. ACMM
    CADQ: Attribute-Consistent Face Cartoonization with Cross-modal Aligned and Deformable Quantization
    Yongjie Hu, Yifan Jiang, Ziyun Li, and 3 more authors
    In ACM International Conference on Multimedia, 2025
  6. AAAI
    QuARF: Quality-Adaptive Receptive Fields for Degraded Image Perception
    Fei Gao, Ying Zhou, Ziyun Li, and 5 more authors
    AAAI Conference on Artificial Intelligence, 2025
  7. TIP
    S^3OIL: Semi-Supervised SAR-to-Optical Image Translation via Multi-Scale and Cross-Set Matching
    Xi Yang, Haoyuan Shi, Ziyun Li, and 3 more authors
    IEEE Transactions on Image Processing, 2025
  8. ICME
    BI-RADS Boosted Breast Cancer Diagnosis With Masked Pretraining On Imbalanced Ultrasound Data
    Xueqian Pang, Ziyun Li, Junhui Lv, and 3 more authors
    In IEEE International Conference on Multimedia and Expo (ICME), 2025
  9. arXiv
    FlowTS: Time Series Generation via Rectified Flow
    Yang Hu, Xiao Wang, Zezhen Ding, Lirong Wu, Huatian Zhang, Stan Z. Li, Sheng Wang, Jiheng Zhang, Ziyun Li, and Tianlong Chen
    arXiv preprint, 2025
    Co-corresponding author

2024

  1. IJCNN
    Otem-IGCD: An Optimal Transport-based EM Framework for Imbalanced Generalized Category Discovery
    Ziyun Li, Ben Dai, Christoph Meinel, and 1 more author
    In International Joint Conference on Neural Networks (IJCNN), 2024
    Oral

2023

  1. TMLR
    Supervised Knowledge May Hurt Novel Class Discovery Performance
    Ziyun Li, Jona Otholt, Ben Dai, and 3 more authors
    Transactions on Machine Learning Research, 2023
    CoLLAs Reward
  2. ICCV
    Generalized Categories Discovery for Long-tailed Recognition
    Ziyun Li, Christoph Meinel, and Haojin Yang
    In Workshop on Out-of-distribution Generalization in Computer Vision, ICCV, 2023
  3. IJCNN
    SMKD: Selective Mutual Knowledge Distillation
    Ziyun Li, Christoph Meinel, Haojin Yang, and 4 more authors
    In International Joint Conference on Neural Networks (IJCNN), 2023
    Oral
  4. CVPR
    ImbaGCD: Imbalanced Generalized Category Discovery
    Ziyun Li, Ben Dai, Christoph Meinel, and 1 more author
    In CVPR Workshop, 2023
    Spotlight

2022

  1. NeurIPS
    A Closer Look at Novel Class Discovery from the Labeled Set
    Ziyun Li, Jona Otholt, Ben Dai, and 3 more authors
    In Workshop on Distribution Shifts, NeurIPS, 2022
  2. NeurIPS
    Not All Knowledge Is Created Equal: Mutual Distillation of Confident Knowledge
    Ziyun Li, Christoph Meinel, Haojin Yang, and 4 more authors
    In Workshop on Trustworthy and Socially Responsible Machine Learning, NeurIPS, 2022

2019

  1. Neurocomputing
    Style-adaptive Photo Aesthetic Rating via Multi-task Learning
    Fei Gao, Ziyun Li, Jun Yu, and 2 more authors
    Neurocomputing, 2019

2017

  1. ICME
    Convolutional neural networks for intestinal hemorrhage detection in wireless capsule endoscopy images
    Panpeng Li, Ziyun Li, Fei Gao, and 2 more authors
    In IEEE International Conference on Multimedia and Expo (ICME), 2017