KOSTAS PAPAFITSOROS
Welcome to my personal webpage!
I am a Lecturer in Mathematical Data Science at the School of Mathematical Sciences, Queen Mary University of London.
My research lies in the areas of Mathematical Imaging, Inverse Problems, Variational Regularisation Methods, Optimisation and Deep Learning. I have a second research direction which is the result of my long-term involvement in environmental conservation with special focus on sea turtles, see the corresponding section of this website.
You can contact me at: k (dot) papafitsoros (at) qmul (dot) ac (dot) uk
News
- 05/2026: I am giving a talk at the workshop Mathematical Methods in MRI-SFB MR-DYNAMO as well as a seminar talk at the University of Graz.
- 04/2026: New preprint (perspective) Centering ecological goals in automated identification of individual animals.
- 03/2026: New published conference paper LifeCLEF 2026 Teaser: AI Challenges for Biodiversity Understanding and Ecosystem Management, at Advances in Information Retrieval, ECIR 2026, Lecture Notes in Computer Science.
- 03/2026: Short article at the at the 2026 SWOT report on "FAQs about sea turtles: Do sea turtles communicate?", with Damien Chevalier.
- 03/2026: We wrote a short communication article for the Divers for the Environment magazine by the Emirates Diving Association, featuring our work with sea turtles and AI-driven photo-identification at the Sharja Marine Science Research Centre, University of Khorfakkan. Together with Lukáš Adam, Jon Martin, Brendan Godley, Stephen Widdicombe, Henrik Stahl.
- 03/2026: New preprint Degradation-based augmented training for robust individual animal re-identification, with Thanos Polychronou, Lukáš Adam, Viktor Penchev.
- 02/2026: New preprint Learning spatially adaptive sparsity level maps for arbitrary convolutional dictionaries, with Joshua Schulz, David Schote, Christoph Kolbitsch and Andreas Kofler. The paper has been accepted at the Proceedings of the 2026 IEEE International Conference on Image Processing.
- 02/2026: New preprint Split, Skip and Play: Variance-reduced ProxSkip for tomography reconstruction is extremely fast, with Evangelos Papoutsellis and Zeljko Kereta. The paper has been accepted at the Proceedings of the 2026 IEEE International Conference on Image Processing.
- 01/2026: Together with TurtleWatch Egypt 2.0, we launched the second version of the TurtleTurtle wildlife photo-ID game: TurtleTurtle Red Sea! You can learn more about these games at this recent article of QMUL's Centre for Public Engagement.
- 12/2025: I was happy to receive a 2025 Impact Research Award from the Faculty of Science and Engineering, Queen Mary University of London.
News Archive
