It was inspiring to see strong research presentations, thoughtful discussions, and excellent engagement from our students and academic community. A special congratulations to our two PhD researchers Mr Nour Badr and Mr Jack Waller  who were awarded prizes in the oral presentation category, achieving First Prize and Third Prize. Congratulations to you both on this fantastic achievement! 

Nour recently presented his paper, "FRLD-DF: Frequency Ring-Guided LoRA Adaptation of DINOv2 Vision Transformer for Generalizable Deepfake Detection", at the 2026 International Conference on Automatic Face and Gesture Recognition (FG 2026) in Kyoto, Japan.

This work introduces a novel framework that combines parameter-efficient adaptation of DINOv2 with structured frequency-domain modeling for deepfake detection. The proposed approach demonstrates strong generalization capabilities across multiple unseen datasets, achieving state-of-the-art cross-dataset performance while maintaining a low number of trainable parameters. 

Daniel presented his work, “Solving Larger Travelling Salesman Problem Networks with a Penalty-Free Variational Quantum Algorithm,” as a poster at QCTiP 2026. His work explores new approaches to solving complex optimisation problems using variational quantum algorithms, contributing to ongoing research at the intersection of quantum computing and combinatorial optimisation. 

Dr. Xing Liang was honored to serve as a PhD viva committee member for four doctoral defenses in the Doctorate in Physics and Astronomy at the University of Florence. The defenses reflected a high standard of research, and it was a pleasure to contribute to the evaluation process and academic discussions. 

As the year came to a close, our Quantum AI group came together for a final discussion to share ideas and reflections. We looked back on several key milestones from the year and considered how they shaped our collective progress. The session gently reaffirmed the importance of collaboration in moving our work forward. As we enter the new year, we do so with a sense of clarity, steady curiosity, and quiet momentum. 

Nour presented his paper on “FlashF5-MaViT: A Fast Five Frequency Mamba with CDC–FastViT Architecture for Deepfake Detection". This work presents a lightweight and efficient deepfake detection framework that combines multi-frequency spectral analysis with a fast hybrid spatial backbone. Despite having only 4.21M parameters, FlashF5-MaViT outperforms many state-of-the-art models, including those with up to 86M parameters, on challenging cross-dataset benchmarks like Celeb-DF, WildDeepfake, and DFD, while also having 29x faster inference than many baselines. 

The QAI group attended the 2025 National Quantum Technologies Showcase in London, organised by Innovate UK with EPSRC and the UK National Quantum Technologies Programme, to learn about recent developments in quantum science and technology. 

We’re delighted to share that Mr. Daniel Goldsmith spoke on “What is Quantum Computing” at the Year of Quantum event, hosted by Digital Catapult on Wednesday, October 1, 2025. 

The event featured inspiring talks and engaging discussions on the fundamentals of quantum computing, its groundbreaking applications, and the promising future of quantum technologies.

N. E. A. Badr, J. -C. Nebel, D. Greenhill and X. Liang, "WaViT-CDC: Wavelet Vision Transformer With Central Difference Convolutions for Spatial-Frequency Deepfake Detection," in IEEE Open Journal of Signal Processing, vol. 6, pp. 621-630, 2025, doi: 10.1109/OJSP.2025.3571679.  

This research focuses on building more efficient and explainable intrusion detection systems (IDS) for cloud environments where we proposed a lightweight IDS leveraging FastViT, combined with advanced feature selection to significantly improve the detection of minority intrusion attack types such as Web and Infiltration. 

Sponsored by the Turing Grant and hosted by Prof. Filippo Caruso,  Jack undertook a one-month PhD exchange visit with the Quantum Artificial Intelligence (QAI) research group at the University of Florence, Italy, in May 2025. During this visit, he expanded his professional network and established meaningful collaborations with researchers dedicated to advancing the field of quantum machine learning. 

Nikhil Sushil Muneshwar, Xing Liang, and Gordon Hunter, "Football Analysis System using Computer Vision and Machine Learning", In: 12th MathSport International Conference 2025, 4-6 Jun 2025: Luxembourg, Luxembourg

A highly insightful and engaging seminar on Quantum AI was delivered by our academic visitor, Prof. Filippo Caruso, leader of the Quantum AI (Q-AI) research group from the University of Florence, Italy. 

Our researcher Daniel has attended the workshop on QML in Vienna.  It was an amazing session with very interesting lab tours.