Dr Zeyu Fu

I am a Lecturer (Assistant Professor) in Computer Vision and Machine Learning at the Department of Computer Science, University of Exeter, where I lead the Multimodal Intelligence Lab. We aim to investigate how Multimodal AI along with other disciplines helps develop a greener, healthier, and fairer society underpinned by UoE’s 2030 strategy. We’re specifically focusing on the core development of computer vision and machine learning and its applications in healthcare, environment, and social science.

Before that, I was a postdoctoral researcher at the Department of Engineering Science, University of Oxford, and was a member of the Oxford Biomedical Image Analysis (BioMedIA) group, advised by Prof Alison Noble and Dr Michael Suttie. I worked on an NIH-funded project which is in conjunction with the Collaborative Initiative on Fetal Alcohol Spectrum Disorders (CIFASD), to develop a fully automated, objective evaluation of facial features associated with FASD, utilizing 3D surface modelling, deep learning and shape analysis. I also worked on an ERC-funded project named Perception Ultrasound by Learning Sonographic Experience (PULSE), which aims to develop multi-modal machine learning and computer vision systems that can model the sonographer’s expertise to reduce the need for highly trained ultrasound operators.

Before Oxford I was a research assistant at the School of Engineering, Newcastle University, advised by Prof Satnam Dlay. I worked on an MRC-CiC sponsored project to apply cutting-edge non-contact ocular imaging, and machine/deep learning techniques to find a means of improving diagnosis of the most common medical complication of pregnancy, pre-eclamptic toxaemia. At the same institution, I obtained a PhD degree in signal processing and machine learning and was a member of Signal Processing and AI research group, advised by Prof Jonathon Chambers and Dr Mohsen Naqvi. I worked on a DSTL & EPSRC funded project ‘Signal Processing Solutions for the Networked Battlespace’ and was part of the LSSCN Consortium of University Defence Research Collaboration (UDRC), where I developed novel data association algorithms for multiple human tracking in video.

Highly self-motivated UG/MSc/PhD students, research fellows, and visiting students/scholars are welcome to join my team. Please drop me an email along with your CV and research plan if you are interested in working with me.

News

• [April 2025] I have been invited to serve as a Guest Editor for a forthcoming Special Issue in Bioengineering, focused on AI-driven imaging and analysis for biomedical applications. (Deadline: 1 October 2025)  
• [Mar 2025] Congrats to Shuaiyu and the team for a paper about Mamba-based oil spill detection getting accepted in ICLR 2025-ML4RS. (Oral Presentation)
• [Feb 2025] A joint PhD opportunity with the University of Exeter & Université Paris-Saclay! The project is to work on Generative Multimodal Fusion & Uncertainty Quantification for Environmental Monitoring. Please get in touch with your CV and a research plan before the deadline: 30 Mar 2025.
• [Jan 2025] PhD Studentships with EPSRC DLA for September 2025 Entry! Join my Multimodal Intelligence Lab to work on a project titled "Beyond Texts: Harnessing Multimodal AI to Combat Online Harmful Content in Videos."
• [Dec 2024] Congrats to Fu Wang and the team for a paper about evaluating semantic robustness in Bird's Eye View Detection getting accepted in AAAI 2025.
• [Nov 2024] We're hiring Winter/Summer Research Interns (50% FTE) to join our Multimodal Intelligence Lab! Work on multimodal ML, video analysis, multimedia computing, and ML for medical imaging or remote sensing. UK-based & right to work required. Email your CV + research statement to apply.
• [Oct 2024] Exciting Opportunities to join our Multimodal Intelligence Lab at the University of Exeter: Exeter-CSC PhD Scholarships are open for applications. please get in touch with your CV and research plans before the deadline: 2 Dec 2024.
• [Sep 2024] A paper titled "Pose-Oriented Scene-Adaptive Matching for Abnormal Event Detection" co-authored with Newcastle University is accepted for publication in Neurocomputing.
• [Aug 2024] A paper titled "Position and Orientation-Aware One-Shot Learning for Medical Action Recognition from Signal Data" co-authored with Newcastle University is accepted for publication in IEEE Transactions on Multimedia
• [Feb 2024] We are seeking a Postdoctoral Research Fellow at the University of Exeter (in collaboration with Dr Jianbo Jiao from the University of Birmingham) to work on multi-modal learning (visual, text and audio) for video understanding.
• [Jan 2024] A paper titled "Automating the Human Action of First-trimester Biometry Measurement from Real-world Freehand Ultrasound" from the PULSE project is accepted for publication in Ultrasound in Medicine & Biology.
• [Oct 2023] Congrats to Fu Wang (co-supervised with Dr Wenjie Ruan) for his work on assessing the reliability of machine learning optimization getting accepted for presentation in the NeurIPS 2023 workshop (Optimization for Machine Learning).
• [July 2023] A paper titled "Abnormal event detection for video surveillance using an enhanced two stream fusion method" co-authored with Newcastle University is accepted for publication in Neurocomputing.
• [May 2023] Congrats to Fu Wang (co-supervised with Dr Wenjie Ruan) for his first MICCAI paper Self-adaptive Adversarial Training for Robust Medical Segmentation getting early accepted (top 14%)!
• [May 2023] A paper titled "A Scene-Adaptive Framework for Pose-Oriented Abnormal Event Detection" co-authored with Newcastle University is accepted for publication in EUSIPCO 2023. (Oral presentation)
• [Feb 2023] A paper titled "One-shot Medical Action Recognition with A Cross-Attention Mechanism and Dynamic Time Warping" co-authored with Newcastle University is accepted for publication in IEEE ICASSP 2023.
• [Jan 2023] A paper titled "Towards Multi-sweep Ultrasound Video Understanding: Application in Detection of Breech Position using Statistical Priors" from the CALOPUS project is accepted for publication in IEEE ISBI 2023. (Best Runner-Up Oral Presentation)
• [Nov 2022] A paper titled "A Machine Learning Method for Automated Description and Workflow Analysis of First Trimester Ultrasound Scans" from the PULSE project is accepted for publication in IEEE Transactions on Medical Imaging.
• [Oct 2022] Our paper titled "Anatomy-Aware Contrastive Representation Learning for Fetal Ultrasound" wins the Best Paper Award at ECCV2022-MCV and is featured in Best of ECCV of Computer Vision News!