$ whoami

Fuyan Zhang

Efficient deep learning: pruning, knowledge distillation, input-adaptive computation. Now heading toward 3D human motion generation — generative motion models fast enough to run anywhere.

edu:MSc High Performance Computing with Data Science, University of Edinburgh

seek:PhD position · 2027 entry · efficient generative models / 3D human motion

mail:F.Zhang-42@sms.ed.ac.uk

link: Google Scholar·GitHub·Edinburgh, Scotland

$ ls publications/ # selected · 3 of 7 papers

Showing 3 selected publications.

Dyna-Pruner framework jointly producing data and model masks for sparse inference — Fig. 1 · Joint data–model pruning framework

ICME 2026 · SPOTLIGHT★ first author◆ selected

DYNA-PRUNER: Input-Adaptive Data-Model Co-Pruning for Efficient and Scalable Spatio-Temporal Media Prediction

F. Zhang, Y. Li, Q. Xu, Y. Tian, E.S.L. Ho

Preprint

@misc{zhang2026dynapruner,
  title         = {DYNA-PRUNER: Input-Adaptive Data-Model Co-Pruning for Efficient and Scalable Spatio-Temporal Media Prediction},
  author        = {Zhang, Fuyan and Li, Yuqi and Xu, Qing and Tian, Yingli and Ho, Edmond S. L.},
  year          = {2026},
  eprint        = {2606.15346},
  archivePrefix = {arXiv},
  primaryClass  = {cs.CV},
  doi           = {10.48550/arXiv.2606.15346},
  url           = {https://arxiv.org/abs/2606.15346},
  note          = {ICME 2026 Spotlight Paper}
}

AdvVPT framework for fair medical image segmentation with adversarial visual prompts — Fig. 1 · Adversarial visual prompt tuning

INFORMATION SCIENCES · JCR-Q1◆ selected

Achieving Fair Medical Image Segmentation in Foundation Models with Adversarial Visual Prompt Tuning

Y. Li, Y. Li, K. Zhang, F. Zhang, C. Yang, Z. Guo, W. Ding, T. Huang

PDF

@article{li2025advvpt,
  title     = {Achieving Fair Medical Image Segmentation in Foundation Models with Adversarial Visual Prompt Tuning},
  author    = {Li, Yuqi and Li, Yanli and Zhang, Kai and Zhang, Fuyan and Yang, Chuanguang and Guo, Zhongliang and Ding, Weiping and Huang, Tingwen},
  journal   = {Information Sciences},
  volume    = {720},
  pages     = {122501},
  year      = {2025},
  month     = dec,
  publisher = {Elsevier},
  doi       = {10.1016/j.ins.2025.122501},
  url       = {https://doi.org/10.1016/j.ins.2025.122501}
}

Reinforcement learning pipeline for adaptive K-space sampling and medical image segmentation — Fig. 1 · RL-driven K-space sampling pipeline

IEEE TETCI · JCR-Q1◆ selected

Efficient Medical Image Segmentation via Reinforcement Learning-Driven K-Space Sampling

Y. Li, H. Zeng, F. Zhang, C. Yang, Y. Li, W. Ding

PDF

@article{li2026kspace,
  title     = {Efficient Medical Image Segmentation via Reinforcement Learning-Driven K-Space Sampling},
  author    = {Li, Yuqi and Zeng, Hansheng and Zhang, Fuyan and Yang, Chuanguang and Li, Yanli and Ding, Weiping},
  journal   = {IEEE Transactions on Emerging Topics in Computational Intelligence},
  volume    = {10},
  number    = {3},
  pages     = {2596--2609},
  year      = {2026},
  month     = jun,
  publisher = {IEEE},
  doi       = {10.1109/TETCI.2025.3621221},
  url       = {https://doi.org/10.1109/TETCI.2025.3621221}
}

Comparison motivating multi-band amplitude mixup and domain-specific mask image reconstruction — Fig. 1 · Motivation for MRAM and DMIR

CVIU · JCR-Q1

Towards Robust Medical Image Segmentation: Spectro-Spatial Domain Generalization with MRAM and DMIR

Y. Li, J. Dong, H. Zeng, F. Zhang, Z. Dong, C. Yang, Y. Tian

PDF

@article{li2026mram,
  title     = {Towards Robust Medical Image Segmentation: Spectro-Spatial Domain Generalization with MRAM and DMIR},
  author    = {Li, Yuqi and Dong, Junhao and Zeng, Hansheng and Zhang, Fuyan and Dong, Zeyu and Yang, Chuanguang and Tian, Yingli},
  journal   = {Computer Vision and Image Understanding},
  volume    = {270},
  pages     = {104822},
  year      = {2026},
  month     = aug,
  publisher = {Elsevier},
  doi       = {10.1016/j.cviu.2026.104822},
  url       = {https://doi.org/10.1016/j.cviu.2026.104822}
}

PDD teacher-student knowledge distillation pipeline and dynamic mask mechanism — Fig. 2 · PDD pipeline and dynamic mask

COGNITIVE COMPUTATION · JCR-Q1

PDD: Pruning Neural Networks During Knowledge Distillation

D. Xi, W. Yang, F. Zhang, Y. Zhou, Z. Yu, Z. Qiu, B. Zhao, Z. Dong, L. Huang, C. Yang

PDF

@article{xi2024pdd,
  title     = {PDD: Pruning Neural Networks During Knowledge Distillation},
  author    = {Xi, Dan and Yang, Wenjie and Zhang, Fuyan and Zhou, Yihang and Yu, Zhuojun and Qiu, Zhen and Zhao, Boyuan and Dong, Zeyu and Huang, Libo and Yang, Chuanguang},
  journal   = {Cognitive Computation},
  volume    = {16},
  number    = {6},
  pages     = {3457--3467},
  year      = {2024},
  month     = aug,
  publisher = {Springer},
  doi       = {10.1007/s12559-024-10350-9},
  url       = {https://doi.org/10.1007/s12559-024-10350-9}
}

CPG method showing DFS-guided channel grouping and L2-norm pruning — Fig. 1 · DFS-guided grouping and pruning

ICONIP 2024 · CCF-C / CORE-A

CPG: Channel Pruning with DFS Guided Grouping for Efficient Medical Image Segmentation

X. Yan, F. Zhang, Y. Li, Z. Qiu, B. Zhao, J. Li, H. Zeng, Y. Chen, C. Yang

PDF

@incollection{yan2025cpg,
  title     = {CPG: Channel Pruning with DFS Guided Grouping for Efficient Medical Image Segmentation},
  author    = {Yan, Xilin and Zhang, Fuyan and Li, Yuqi and Qiu, Zhen and Zhao, Boyuan and Li, Jiarui and Zeng, Hansheng and Chen, Yihao and Yang, Chuanguang},
  booktitle = {Neural Information Processing},
  pages     = {284--297},
  year      = {2025},
  publisher = {Springer Nature Singapore},
  isbn      = {978-981-96-6585-3},
  doi       = {10.1007/978-981-96-6585-3_20},
  url       = {https://doi.org/10.1007/978-981-96-6585-3_20}
}

Knowledge-distillation network with teacher and MobileNetV3 student for automatic meter reading — Fig. 1 · KKN teacher–student architecture

KSEM 2024 · CCF-C / CORE-B

Automatic Meter Pointer Reading Based on Knowledge Distillation

S. Rong, W. Yang, F. Zhang, Y. Xiang, H. Wang, Y. Jiang

PDF

@incollection{rong2024meter,
  title     = {Automatic Meter Pointer Reading Based on Knowledge Distillation},
  author    = {Rong, Sun and Yang, Wenjie and Zhang, Fuyan and Xiang, Yanzhuo and Wang, Hengxi and Jiang, Yuncheng},
  booktitle = {Knowledge Science, Engineering and Management},
  pages     = {376--392},
  year      = {2024},
  publisher = {Springer Nature Singapore},
  isbn      = {978-981-97-5492-2},
  doi       = {10.1007/978-981-97-5492-2_29},
  url       = {https://doi.org/10.1007/978-981-97-5492-2_29}
}

$ cat education.log

2025.09 — 2026.09

University of Edinburgh

MSc in High Performance Computing with Data Science

2022.09 — 2025.06

University of Glasgow

BSc (Honours) in Computing Science

$ cat experience.log

2025.07 — 2025.09

Research Intern · University of Glasgow

School of Computing Science · Supervisor: Dr. Edmond S.L. Ho — dynamic data–model co-pruning for spatio-temporal prediction (→ Dyna-Pruner, ICME 2026 Spotlight)

2024.06 — 2024.08

Research Assistant · Institute of Computing Technology, CAS

Beijing · Supervisor: Dr. Chuanguang Yang — model pruning and knowledge distillation across medical segmentation and detection (→ 4 co-authored papers)