Dr Pengfei Fan

Profile

Dr Pengfei Fan is a Lecturer in Data Science and AI at Queen Mary University of London (QMUL) and a Fellow of the Digital Environment Research Institute (DERI). His interdisciplinary research operates at the nexus of Artificial Intelligence, Optics, and Neural Engineering, with a primary vision to pioneer next-generation, high-bandwidth Brain-Computer Interfaces (BCIs).

 

By leveraging computational photonics and Deep Generative AI, he develops scalable, AI-enhanced neural interfaces that transform flexible multimode fibres into ultra-thin, minimally invasive optical probes. This technology enables real-time neural decoding, high-fidelity deep-brain mapping, and closed-loop neuromodulation, bypassing the limitations of traditional invasive electrodes. Recognised for his innovation, Dr Fan directed a multi-institutional team (QMUL, Imperial and Oxford) to secure 1st Place at the 2026 London Neurotech Hackathon (Bluesky track) for their work on hybrid neural interfaces.

 

Dr Fan earned his PhD from QMUL, where his research focused on deep learning for imaging through dynamic scattering media. Before returning to QMUL as faculty, he held academic positions at Xi'an Jiaotong-Liverpool University, Nanjing University of Science and Technology, and Tsinghua University. He is a Fellow of the Higher Education Academy (FHEA) and actively contributes to AI-driven education initiatives.

Teaching

Current Teaching Modules

• QHE4102: Introduction to Artificial Intelligence
• QHE5002: Sensors, Wearables and Brain-Computer Interfaces
• EBU6232: AI in Games
• QHM6706: Undergraduate Final Year Project (Supervisor)
• SPC720P: Artificial Intelligence and Machine Learning in Science MSc Project (Supervisor)

 

Academic Leadership

• Programme Director: Information and Computational Science (UG)
• Deputy Programme Director: Intelligent Biomedical Engineering (UG)
• Committee Member: Joint Teaching and Learning Centre (JTLC), QMUL
• Affiliate Member: Centre for Excellence in AI in Education, QMUL

 

Teaching Excellence & Recognition

• Fellow of the Higher Education Academy (FHEA)
• Principal Investigator: Awarded £20,000 from the QMUL President and Principal’s Fund for Educational Excellence
• University Research-led Teaching Award

Research

Research Interests:

Dr Pengfei Fan’s research resides at the innovative intersection of Artificial Intelligence, Computational Photonics, and Neural Engineering. He leverages advanced Data Science and AI methodologies to decode complex optical information, aiming to pioneer next-generation, high-bandwidth Brain-Computer Interfaces (BCIs) and minimally invasive diagnostic tools.

 

Core Research Interests

• AI-Enhanced Optical BCI: Developing scalable, ultra-thin neural interfaces based on multimode fibres (MMF) that use AI for real-time neural decoding and deep-brain signal acquisition.
• Generative AI for Computational Imaging: Utilising Deep Generative Models (e.g., Diffusion Models, GANs) to achieve high-fidelity image reconstruction through dynamic scattering media.
• Physics-Informed Machine Learning: Integrating optical priors into neural networks to solve complex inverse problems in variable biological environments.

 

Collaborative Network

Dr Fan maintains an extensive network of interdisciplinary collaborations to translate AI and optical breakthroughs into clinical and engineering impact:

• Internal (QMUL):

Prof Lei Su (School of Engineering and Materials Science)

Prof Valentina Donzella (School of Engineering and Materials Science)

Dr James Timmons (William Harvey Research Institute

Prof Paul Chapple (William Harvey Research Institute)

 

• External:

Prof Dario Farina (Imperial College London)

Dr Mathieu Bourdenx (University College London / UK Dementia Research Institute)

Prof Chao Zuo (Nanjing University of Science and Technology)

 

Research Opportunities

PhD applications through CSC, CONACYT, and HEC are welcome. Happy to support potential applications for Research Training Fellowships. Dr Fan is particularly interested in students passionate about AI in Healthcare and Neurotechnology.

Examples of research funding:

Pengfei's research is supported by a diverse portfolio of national and international funding bodies, including the EPSRC, the Royal Society, the NSFC, and the Jiangsu Science and Technology Programme, among others.

 

Research Grants (Principal Investigator)

• Fast High-Resolution Imaging through Flexible Multimode Fibers Based on Back-Reflection Guided Conditional Diffusion Models, NSFC, Young Scientists Fund.
• Long-Term High-Fidelity Image Transmission through Multimode Fibers Based on Temporal-Spatial Dynamics-Informed Diffusion Models, Jiangsu Science and Technology Programme, Young Scientists Fund.
• Learning-Based Methods for Single Image Restoration and Enhancement, University of Liverpool Research Development Fund.

 

Research Grants (Co-Investigator / Contributor)

• Key Technology Development for Small-Sample Industrial Vision Inspection, Suzhou Science and Technology Programme (Co-PI).
• Flexible Single-Optical-Fibre Endoscope, EPSRC and Royal Society, Research Contributor.

 

Teaching & Educational Grants

• Enhancing Data Science Education through Competitive-Based Learning and AI-Driven Assessment, QMUL President and Principal’s Fund for Educational Excellence (PI).
• Enhancing Big Data Analytics Teaching through Competition-Based Learning Using Kaggle Platform, University of Liverpool Teaching Development Fund (PI).
• Chatbot-Powered Learning for Sustainable Education in Programming, Advance HE Global Impact Grant (Co-PI).

Publications

Full list of publications can be found on Google Scholar.

Fan, P., Wang, Y., Ruddlesden, M., Zuo, C., & Su, L. (2024, May). Enhanced Light Control in Transmission and Reflection through a Dynamically Deformed Multimode Fiber with Deep Learning. In CLEO: Applications and Technology(pp. AF1B-2). Optica Publishing Group.

Fan, P., Wang, Y., Ruddlesden, M., Wang, X., Thaha, M.A., Sun, J., Zuo, C. and Su, L., 2022. Deep learning enabled scalable calibration of a dynamically deformed multimode fiber. Advanced Photonics Research, 3(10), p.2100304.

Zuo, C., Qian, J., Feng, S., Yin, W., Li, Y., Fan, P., ... & Chen, Q. (2022). Deep learning in optical metrology: a review. Light: Science & Applications, 11(1), 1-54.

Fan, P., Ruddlesden, M., Wang, Y., Zhao, L., Lu, C., & Su, L. (2021). Learning enabled continuous transmission of spatially distributed information through multimode fibers. Laser & Photonics Reviews, 15(4), 2000348.

Fan, P., Zhao, T., & Su, L. (2019). Deep learning the high variability and randomness inside multimode fibers. Optics express, 27(15), 20241-20258.

JavaException: java.lang.IllegalArgumentException: Illegal character in path at index 66: https://researchpublications.its.qmul.ac.uk/publications/data/Full list of publications can be found on Google Scholar.\n \n\nFan, P., Wang, Y., Ruddlesden, M., Zuo, C., & Su, L. (2024, May). Enhanced Light Control in Transmission and Reflection through a Dynamically Deformed Multimode Fiber with Deep Learning. In CLEO: Applications and Technology(pp. AF1B-2). Optica Publishing Group.\n\n \n\nFan, P., Wang, Y., Ruddlesden, M., Wang, X., Thaha, M.A., Sun, J., Zuo, C. and Su, L., 2022. Deep learning enabled scalable calibration of a dynamically deformed multimode fiber. Advanced Photonics Research, 3(10), p.2100304.\n\n \n\nZuo, C., Qian, J., Feng, S., Yin, W., Li, Y., Fan, P., ... & Chen, Q. (2022). Deep learning in optical metrology: a review. Light: Science & Applications, 11(1), 1-54.\n\n \n\nFan, P., Ruddlesden, M., Wang, Y., Zhao, L., Lu, C., & Su, L. (2021). Learning enabled continuous transmission of spatially distributed information through multimode fibers. Laser & Photonics Reviews, 15(4), 2000348.\n\n \n\n5. Fan, P., Zhao, T., & Su, L. (2019). Deep learning the high variability and randomness inside multimode fibers. Optics express, 27(15), 20241-20258.\n\n \nWang, M., Tang, X., Zhai, L., & Fan, P. \"Dynamic Weighting Diffusion Model with Multi-Scale Feature Fusion for Low-Light Image Enhancement.\" 2026 7th International Conference Computer Vision and Computational Intelligence (CVCI 2026), 2026. \n \nLiu, J., Huo, J., & Fan, P. \"Content-aware Low-light Image Enhancement with User-defined Variable Illumination Using Diffusion Model.\" 2025 8th International Conference on Machine Learning and Machine Intelligence (MLMI 2025), 2025. \n \nLiu, M., Yao, R., Wu, M., & Fan, P. \"FourierDiff: Image Denoising Using Improved Denoising Diffusion Probabilistic Models via Fast Fourier Convolution.\" The 35th British Machine Vision Conference (BMVC 2024), 2024. \n \nFan, P., Wang, Y., Ruddlesden, M., Zuo, C., & Su, L. \"Enhanced Light Control in Transmission and Reflection through a Dynamically Deformed Multimode Fiber with Deep Learning.\" CLEO: Applications and Technology, AF1B. 2, 2024. \n \nXia, W., Bi, Y., Cao, Y., Xu, K., & Fan, P. \"Super-resolved image reconstruction by structured illumination microscopy.\" Advanced Optical Imaging Technologies V, PC123160S, 2023. \n \nZuo, C., Qian, J., Feng, S., Yin, W., Li, Y., Fan, P., Han, J., Qian, K., & Chen, Q. \"Deep learning in optical metrology: a review.\" Light: Science & Applications, 11(1), 39, 2022. \n \nFan, P., Wang, Y., Ruddlesden, M., Wang, X., Thaha, M. A., Sun, J., Zuo, C., & Su, L. \"Deep learning enabled scalable calibration of a dynamically deformed multimode fiber.\" Advanced Photonics Research, 3(10), 2100304, 2022. \n \nFan, P., Ruddlesden, M., Wang, Y., Zhao, L., Lu, C., & Su, L. \"Learning enabled continuous transmission of spatially distributed information through multimode fibers.\" Laser & Photonics Reviews, 15(4), 2000348, 2021. \n \nFan, P., Zhao, T., & Su, L. \"Deep learning the high variability and randomness inside multimode fibers.\" Optics Express, 27(15), 20241-20258, 2019..xml