Junchen Fu

Junchen Fu (付俊臣)

Pronunciation: Joon-chen Foo (/ˈdʒuːn tʃən fuː/)

University of Glasgow

PhD Candidate

School of Computing Science, IDA-section

Member of GAIR Lab

Email: j.fu.3@research.gla.ac.uk;

Google Scholar: Google Scholar Profile     Github: Google Scholar Profile     Twitter(X): Google Scholar Profile

About Me

I am a final-year PhD candidate at the University of Glasgow, advised by Prof. Joemon M. Jose, and have also been fortunate to be mentored by Dr. Alexandros Karatzoglou and Dr. Ioannis Arapakis. I was a visiting PhD student at Leiden University, advised by Prof. Zhaochun Ren and Prof. Suzan Verberne. Previously, I worked as a Research Assistant at Westlake University under the supervision of Prof. Fajie Yuan, focusing on modality-based recommender systems. I received my MSc in Computer Science from The Chinese University of Hong Kong, where I graduated with Dean’s List honors. I have authored over 10 papers as first, co-first, or corresponding author in premier AI conferences and journals.

My research focuses on Recommendation-Oriented Multimodal Foundation Models, with the central question of how to transform general-purpose multimodal foundation models into systems that can represent, generate, and personalize multimodal information for recommendation. Specifically, I study how representation and generation foundation models can be adapted into recommendation-oriented systems that better understand users, represent items, and support personalized information ecosystems.

Research Overview
Research framework for recommendation-oriented multimodal foundation models

My work mainly explores this direction from two perspectives:

  1. Recommendation-Oriented Representation

    I study how to adapt multimodal foundation models into recommendation-aligned representations in both continuous and discrete spaces. My work includes efficient continuous multimodal representation adaptation [ Adapter4Rec (WSDM'24), IISAN (SIGIR'24), IISAN-Versa (TKDE'25), CROSSAN (ArXiv'25) ], and discrete Semantic ID learning for generative recommendation [ DIGER (SIGIR'26) ].

  2. Recommendation-Oriented Generation

    I also explore how generative models can support recommender systems from the user and platform sides, including popular content generation on user side [ LLMPopcorn (ICASSP'26) ] and missing modality completion for item understanding on platform side [ MMPCBench (PR'26) ].

I enjoy tackling non-trivial research questions with simple yet effective solutions.
I am seeking to collaborate on a research project with self-motivated MSc students from UofG, particularly those interested in pursuing a PhD. the students who are interested in research towards premier AI conferences/journals are welcome to drop me an email.

Latest News

Oral Presentation/Invited Talk:

Workshop Summary:

Selected Publications

Note: Corresponding-author papers come from MSc/Undergraduate supervision at UofG or other universities.
  1. Junchen Fu, Xuri Ge, Alexandros Karatzoglou, Ioannis Arapakis, Suzan Verberne, Joemon M. Jose, Zhaochun Ren.
    Differentiable Semantic ID for Generative Recommendation. [CODE]
    - Accepted by SIGIR 2026 (CORE Rank A*, CCF-A)
  2. Junchen Fu, Wenhao Deng, Kaiwen Zheng, Ioannis Arapakis, Yu Ye, Yongxin Ni, Joemon M. Jose, Xuri Ge.
    Benchmarking Multimodal Large Language Models for Missing Modality Completion in Product Catalogues.
    - Accepted by PR 2026 (CORE Rank A*, CCF-B, JCR Q1, IF=7.6)
  3. Junchen Fu, Xuri Ge, Kaiwen Zheng, Alexandros Karatzoglou, Ioannis Arapakis, Xin Xin, Yongxin Ni, Joemon M. Jose.
    LLMPopcorn: Exploring LLMs as Assistants for Popular Micro-video Generation. [CODE]
    - ICASSP 2026 (CCF-B)
  4. Kaiwen Zheng, Junchen Fu, Songpei Xu, Yaoqing He, Joemon M.Jose, Han Hu, Xuri Ge
    Focal-RegionFace: Generating Fine-Grained Multi-attribute Descriptions for Arbitrarily Selected Face Focal Regions.
    - Accepted by ICMR 2026 (CCF-B)
  5. Yu Ye, Junchen Fu# (corresponding author), Yu Song, Kaiwen Zheng, Joemon M. Jose.
    A Reproducibility Study of Multimodal Embeddings for Recommender Systems. [CODE]
    - Accepted by IJMIR 2026 (JCR Q2, IF=2.9), Extension of ECIR2026
  6. Yu Ye, Junchen Fu# (corresponding author), Yu Song, Kaiwen Zheng, Joemon M. Jose.
    Are Multimodal Embeddings Truly Beneficial for Recommendation? A Deep Dive into Whole vs. Individual Modalities. [CODE]
    - ECIR 2026 (CORE Rank A)
  7. Hui Ye, Xuri Ge, Junqi Wang, Junchen Fu, Xin Xin, Jiao Xue, Yao Chen, Pengjie Ren, Zhumin Chen.
    Beyond efficient fine-tuning: Efficient hybrid fine-tuning of CLIP models guided by explainable ViT attention. [CODE]
    - Accepted by IPM 2026 (CORE Rank A, CCF-B, JCR Q1, IF=6.9)
  8. Junchen Fu, Xuri Ge, Xin Xin, Alexandros Karatzoglou, Ioannis Arapakis, Kaiwen Zheng, Yongxin Ni, and Joemon M. Jose
    Efficient and Effective Adaptation of Multimodal Foundation Models in Sequential Recommendation. [CODE]
    - TKDE2025 (CORE Rank A*, CCF-A, JCR Q1, IF=10.4)
  9. Zichen Yuan, Lifan Sun, Yucen Zhuang, Yue Wang, Xinyuan Song, Tianqi Xu, Siyuan Li, Junchen Fu, Youhua Li, Sirui Hong, Jiaqi Chen, Joemon M. Jose, Yongxin Ni.
    SOLAR: Serendipity Optimized Language Model Aligned for Recommendation [CODE].
    - EMNLP 2025 (Finding)
  10. Yongxin Ni, Yu Cheng, Xiangyan Liu, Junchen Fu, Youhua Li, Xiangnan He, Yongfeng Zhang, Fajie Yuan.
    A Content-Driven Micro-Video Recommendation Dataset at Scale. [CODE]
    - CIKM2025 (CORE Rank A, CCF-B)
  11. Ziyi Zhuang, Hongji Li, Junchen Fu# (corresponding author), Jiacheng Liu, Joemon M. Jose, Youhua Li, Yongxin Ni.
    Frequency-Decoupled Distillation for Efficient Multimodal Recommendation.
    - CIKM2025 (CORE Rank A, CCF-B)
  12. Yaoqin He*, Junchen Fu*(co-first authors), Kaiwen Zheng, Songpei Xu, Fuhai Chen, Jie Li, Joemon M. Jose, and Xuri Ge.
    Double-Filter: Efficient Fine-tuning of Pre-trained Vision-Language Models via Patch&Layer Filtering.
    - ICML2025 (CORE Rank A*, CCF-A)
  13. Junchen Fu, Yongxin Ni, Joemon M. Jose, Ioannis Arapakis, Kaiwen Zheng, Youhua Li, and Xuri Ge.
    CROSSAN: Towards Efficient and Effective Adaptation of Multiple Multimodal Foundation Models for Sequential Recommendation.
    -Arxiv 2025 (Preprint)
  14. Youhua Li, Ersheng Ni, Yihao Liu, Sibo Xu, Tianyi Xu, Mingxuan Wu, Junchen Fu, Yucheng Zhang, Yuanqi He, Xinyuan Song, Yongxin Ni.
    Bridging NIP and MLM: A Unified Meta-Learning Framework for Sequential Recommendation.
    - TKDD2025 (CCF-B, JCR Q1, IF=4.8)
  15. Kaiwen Zheng, Xuri Ge, Junchen Fu, Jun Peng, Joemon M. Jose.
    Multimodal Representation Learning Techniques for Comprehensive Facial State Analysis.
    - ICME2025 (CORE Rank A, CCF-B)
  16. Hui Han, Siyuan Li, Jiaqi Chen, Yiwen Yuan, Yuling Wu, Yufan Deng, Chak Tou Leong, Hanwen Du, Junchen Fu, Youhua Li, Jie Zhang, Chi Zhang, Li-jia Li, Yongxin Ni.
    Human-Aligned Video Generation Benchmark.
    - CVPR'2025 (CORE Rank A*, CCF-A)
  17. Ziyi Zhuang, Hanwen Du, Hui Han, Youhua Li, Junchen Fu, Joemon M. Jose, Yongxin Ni.
    Bridging the Gap: Teacher-Assisted Wasserstein Knowledge Distillation for Efficient Multi-Modal Recommendation.
    - ACM WWW'2025 (CORE Rank A*, CCF-A)
  18. Zhiyu Liu, Junchen Fu# (corresponding author), Kaiwen Zheng, Joemon M. Jose.
    Exploring Multimodal Pre-trained Models for Speech Emotion Recognition.
    - ACM WWW'2025 Workshop
  19. Hongji Li, Hanwen Du, Youhua li, Junchen Fu, Chunxiao Li, Ziyi Zhuang, Jiakang Li, Yongxin Ni.
    Teach Me How to Denoise: a Universal Framework for Denoising Multi-modal Recommender Systems via Guided Calibration [CODE].
    - ACM WSDM'2025 (CORE Rank A, CCF-B, Best of WSDM)
  20. Xuri Ge*, Junchen Fu*(co-first authors), Fuhai Chen, Shan An, Nicu Sebe, Joemon Jose.
    Towards End-to-End Explainable Facial Action Unit Recognition via Vision-Language Joint Learning.
    - ACM MM'2024 (CORE Rank A*, CCF-A).
  21. Junchen Fu, Xuri Ge, Xin Xin, Alexandros Karatzoglou, Ioannis Arapakis, Jie Wang, Joemon Jose.
    IISAN: Efficiently Adapting Multimodal Representation for Sequential Recommendation with Decoupled PEFT. [CODE]
    - SIGIR'2024 (CORE Rank A*, CCF-A).
  22. Junchen Fu, Fajie Yuan, Yu Song, Zheng Yuan, Mingyue Cheng, Shenghui Cheng, Jiaqi Zhang, Jie Wang, Yunzhu Pan.
    Exploring Adapter-based Transfer Learning for Recommender Systems: Empirical Studies and Practical Insights. [CODE]
    - WSDM'2024 (CORE Rank A, CCF-B).
  23. Jiaqi Zhang, Yu Cheng, Yongxin Ni, Yunzhu Pan, Zheng Yuan, Junchen Fu, Youhua Li, Jie Wang, Fajie Yuan.
    NineRec: A Benchmark Dataset Suite for Evaluating Transferable Recommendation. [CODE]
    - TPAMI2024 (CORE Rank A*, CCF-A, JCR Q1, IF=20.8)
  24. Zheng Yuan, Fajie Yuan, Yu Song, Youhua Li, Junchen Fu, Fei Yang, Yunzhu Pan, and Yongxin Ni.
    Where to Go Next for Recommender Systems? ID- vs. Modality-based Recommender Models Revisited. [CODE]
    - SIGIR'2023 (CORE Rank A*, CCF-A)
  25. Junchen Fu and Zhaohui Qi.
    A TDF-WNSP-WLFM algorithm for product recommendation based on multiple types of implicit user behavior.
    - The Journal of Supercomputing 2022 (CORE Rank B, CCF-C, JCR Q2, IF=2.5)

Work Experience

Education Background

MSc Supervision

Graduate Teaching Assistant

Professional Activities

Honors and Awards