A paper, FairMT: Fairness for Heterogeneous Multi-Task Learning, has been accepted to NeurIPS 2026. Congrats to Guanyu and all co-authors!
University of Exeter · Computer Science
Multimodal
Intelligence Lab
Welcome to the Multimodal Intelligence Lab, where we explore innovative solutions at the intersection of Artificial Intelligence, Machine Learning, and Multimodal Data Processing. Aligned with the University of Exeter’s 2030 strategy, our research investigates how Multimodal AI and related disciplines can contribute to a greener, healthier, and fairer society. Our core focus is the development of computer vision and machine learning techniques, with applications in healthcare, the environment, and social science.
Research interests
Multimedia Understanding and Safety
9 publicationsMultimedia Understanding and Safety
Fusing and reasoning across vision, audio and text to detect and localise hateful content, deepfakes and misinformation in online media, and building the benchmark datasets that measure this.
9 publications · ExploreAI for Healthcare
7 publicationsAI for Healthcare
Surgical video segmentation, medical image classification and radiology report generation, together with learning from brain and physiological signals for clinical and health applications.
7 publications · ExploreAI for Environment
8 publicationsAI for Environment
Monitoring oil spills, marine pollution and land-surface change in satellite and drone imagery, and deep learning for fast, transferable prediction of urban flooding.
8 publications · ExploreVideo and Scene Understanding
5 publicationsVideo and Scene Understanding
Anomaly detection in surveillance video and 3D point clouds, analysis of human pose, behaviour and facial reactions, and the robustness of visual perception models.
5 publications · ExploreTrustworthy AI
3 publicationsTrustworthy AI
Fairness across groups and tasks, learning from imbalanced and noisy data, and reliable, guaranteed evaluation with large language models.
3 publications · ExploreLab updates
News
29 updates · scroll to explore
A paper, MambaMPD: A Mamba-Driven Segmentation Framework for Marine Pollution Detection from Remote Sensing Imagery, has been accepted to GIScience & Remote Sensing. Congrats to Shuaiyu and all co-authors!
Two papers have been accepted to NeurIPS 2026 Workshops. Congrats to Lisa and all co-authors!
The source announcement does not yet provide the paper titles or links; they will be connected here when public metadata becomes available.
A paper, Generalist–specialist transfer learning for cross-city spatial generalization in pluvial flood prediction, has been accepted to Water Research. Congrats to Zhufeng and all co-authors!
A paper, Localize-Then-Decide Guarantees for LLM Judgments, has been accepted to EMNLP 2026 Findings. Congrats to all co-authors!
A paper, CLARA: Clip-Level Multimodal Alignment with VLM-Derived Rationales for Hateful Video Detection, has been accepted to ACM Multimedia 2026. Congrats to Yuchen and all co-authors!
A paper, ReGround-Surg: Reliability-Guided Anchor Grounding for Referring Surgical Video Segmentation, has been accepted to PRCV 2026 (Oral Presentation, 2.8%). Congrats to Jiaxin (UG researcher) and all co-authors!
A main paper, CPR: Chained Perceptual Refinement for Coarse-to-Fine Medical Image Classification, and a workshop paper have been accepted to MICCAI 2026. Congrats to all co-authors!
The source announcement does not yet identify the workshop paper; it will be linked once public metadata becomes available.
Congrats to Shuonan and the team for Reasoning-Aware Multimodal Fusion for Hateful Video Detection getting accepted to Transactions on Machine Learning Research.
A paper, Probing 3D Anomalies via Multi-View Registration and Dual-Residual Analysis, has been accepted to Neurocomputing. Congrats to Yuxing and all co-authors!
A paper, Cross-Modal Progressive Modeling for Neuro-Visual Representation Learning, has been accepted to Neurocomputing. Congrats to Yueming and all co-authors!
Two papers have been accepted to IGARSS 2026. Congrats to Selorm, Zishu and all co-authors!
The source announcement does not yet provide the paper titles or links; they will be connected here when public metadata becomes available.
Two main papers, Confusion-Aware Spectral Regularizer for Long-Tailed Recognition (Oral) and Omni-Fake: Benchmarking Unified Multimodal Social Media Deepfake Detection, along with a findings paper, MReactor: Offline Multiple Appropriate Facial Reaction Generation with Hierarchical Cognitive Disentanglement, have been accepted to CVPR 2026. Congrats to Tianxiao (UG researcher) and all co-authors!
Congrats to Faezeh and the team for a paper about concept-aware radiology report generation, GraphRAG-Rad, getting accepted in the EACL 2026 Student Research Workshop.
Congrats to Shuaiyu, Shuonan and the team for two papers about SAM-based marine pollution detection and training-free hateful video detection, getting accepted in ICASSP 2026.
Congrats to Shuaiyu and the team for a paper, Enhancing Remote Sensing Change Detection via Masked Edge Reconstruction, getting accepted in the WACV 2026 CV4EO Workshop.
Congrats to Qiyue, Tailin and the team for a paper, MultiHateLoc: Towards Temporal Localisation of Multimodal Hate Content in Online Videos, getting accepted at The Web Conference 2026.
Congrats to Zhufeng and the team for Discriminator-Guided Generative Adversarial Networks for Urban Flood Prediction getting accepted in Water Resources Research.
Congrats to Shuaiyu and the team for Enhancing Marine Pollution Detection in Remote Sensing via Self-Supervised Boundary Awareness getting accepted in the BMVC 2025 MVEO Workshop.
Congrats to Yuchen and the team for a paper, DeHate: A Holistic Hateful Video Dataset for Explicit and Implicit Hate Detection, getting accepted in ACM Multimedia 2025 Dataset Track.
Big congrats to Fu Wang for passing his PhD viva with minor corrections. Well done, Dr Fu Wang!
Congrats to Ruby (UG) and the team for Integrating Semantic, Sentiment, and Object-level Cues for Multimodal Video-based Fake News Detection getting accepted in an ACM Multimedia 2025 Workshop.
Congrats to Shuonan (UG) and the team for a paper, Revealing Temporal Label Noise in Multimodal Hateful Video Classification, getting accepted in ACM Multimedia 2025 Workshop.
Congrats to Jiangbei and the team for a paper, Multimodal Hate Detection Using Dual-Stream Graph Neural Networks, getting accepted in BMVC 2025.
Congrats to Snehil and the team for PySimPace v2.0: An Easy-to-Use Simulation Tool with Machine Learning Pipelines for Realistic MRI Motion Artifact Generation being accepted to the ACM Multimedia 2025 Open Source Track and for receiving a US$1,000 travel grant!
Congrats to Zhufeng and the team for Assessment of Rainfall-Driven Urban Surface Water Flood Hazards Using Convolutional Neural Networks getting accepted in the Journal of Flood Risk Management.
Congrats to Ming and the team for a paper, Memory-Augmented SAM2 for Training-Free Surgical Video Segmentation, getting accepted in MICCAI 2025. (Oral Presentation)
Congrats to Shuaiyu and the team for the full version of OSDMamba getting accepted in IEEE Geoscience and Remote Sensing Letters.
Dr Zeyu Fu has been invited to serve as a Guest Editor for a special issue on AI-driven imaging and analysis for biomedical applications.
Selected Funders
The research works at MIL has gratefully received funding from several distinguished and venerable funders.
Selected Collaborators
The lab is fortunate to collaborate with a range of internal and external partners.












