Skip to main content
School of Mathematical Sciences

Dr Kostas Papafitsoros


Lecturer in Mathematical Data Science

Room Number: MB-117


I am a Lecturer in Mathematical Data Science at the School of Mathematical Sciences, Queen Mary University of London. Prior to that (2017-2022), I was a research scientist at the Weierstrass Institute for Applied Analysis and Stochastics, in Berlin, Germany, working in the group Nonsmooth Variational Problems and Operator Equations. During the period September 2015-August 2017, I was an Alexander von Humboldt Postdoctoral Fellow, working initially at the Mathematical Institute of Humboldt University Berlin and later at the Weierstrass Institute. I completed my PhD in 2014 at the University of Cambridge, where I was also a member of the Cambridge Image Analysis group. I stayed in Cambridge six more months after my PhD, as  an EPSRC Doctoral Prize fellow at the Department of Applied Mathematics and Theoretical Physics.

Undergraduate Teaching

  • MTH6101 Introduction to Machine Learning (2022-23, 2023-24)

Postgraduate Teaching

  • Msc Data Analytics, dissertation supervision (2022-23, 2023-24)


Research Interests:

Spatially varying regularisation parameter
Spatially varying regularisation parameter

My main research area is mathematical imaging, in the interface of several areas of applied mathematics, like inverse problems, variational methods, calculus of variations, functional analysis, optimisation, optimal control, numerical analysis and deep learning. I have been particularly involved in the design, analysis and application of nonsmooth energy functionals incorporated in variational regularisation methods tailored for image processing. These functionals are at one hand discontinuity-preserving often stemming from their nonsmoothness, that is, they have the ability to preserve sharp edges in the images and on the other hand, they are adaptive to the specific structure of the given data. I am interested in combining these "classical" methods with modern data-driven and deep learning approaches.


I have a second research direction which is the result of my long-term involvement in environmental conservation with special focus on marine turtles, see the corresponding section of my personal website. I consider several applications of animal photo-identification (identification of individual animals via their unique morphological characteristics e.g. patterns/features), and in the same time consider various ways to enhance and automatise this process via modern imaging techniques. I am also looking at mathematical/statistical aspects of citizen science projects related to wildlife conservation.

In summary, I am interested in:

  • Mathematical imaging
  • Inverse Problems
  • Optimisation
  • Optimal control
  • Calculus of variations
  • Functional analysis
  • Numerical analysis
  • Deep learning and neural networks
  • Wildlife conservation (marine turtles)
  • Animal photo-identification techniques
  • Mathematical/statistical aspects of citizen science 

Examples of research funding:

  • British Chelonia Group Conservation fund: Addressing the post-pandemic rise of the tourist pressure on an important Mediterranean loggerhead sea turtle population: Quantification, public engagement and influencing policy (£2,500)
  • QMUL Impact Acceleration Fund: Engaging citizens in responsible tourism: Integrating wildlife viewing with artificial intelligence for a novel user experience (£18,290). The project led to the creation of the Zakynthos Turtles platform, see Public Engagement section.


Here are a few latest relevant publications of mine. For a complete list of publications, please visit my google scholar profile.

  • Data-driven methods for quantitative imaging, 
    Guozhi Dong, Moritz Flaschel, Michael Hintermüller, Kostas Papafitsoros, Clemens Sirotenko, Karsten Tabelow, (2024), [arXiv preprint]
  • Machine learning for quantitative MR image reconstruction,
    Andreas Kofler, Felix Zimmerman, Kostas Papafitsoros, (2024), [arXiv preprint]
  • WildlifeDatasets: An open-source toolkit for animal re-identification,
    Vojtech Cermak, Lukas Picek, Lukas Adam, Kostas Papafitsoros, Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, (2024), [DOI]
  • SeaTurtleID2022: A long-span dataset for reliable sea turtle re-identification, 
    Lukas Adam, Vojtech Cermak, Kostas Papafitsoros, Lukas Picek, Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, (2024), [DOI]
  • A descent algorithm for the optimal control of ReLU neural network informed PDEs based on approximate directional derivatives,
    Guozhi Dong, Michael Hintermüller, Kostas Papafitsoros, SIAM Journal on Optimization (to appear), (2024), [arXiv preprint]
  • Learning regularization parameter-maps for variational image reconstruction using deep neural networks and algorithm unrolling,
    Andreas Kofler, Fabian Altekrüger, Fatima Antarou Ba, Christoph Kolbitsch, Evangelos Papoutsellis, David Schote, Clemens Sirotenko, Felix Frederik Zimmermann, Kostas Papafitsoros, SIAM Journal on Imaging Sciences, (2023), [DOI]
  • Unrolled three-operator splitting for parameter-map learning in low dose X-ray CT reconstruction,
    Andreas Kofler, Fabian Altekrüger, Fatima Antarou Ba, Christoph Kolbitsch, Evangelos Papoutsellis, David Schote, Clemens Sirotenko, Felix Frederik Zimmermann, Kostas Papafitsoros, Fully 3D, (2023), [arXiv preprint]
  • SeaTurtleID: A novel long-span dataset highlighting the importance of timestamps in wildlife re-identification,
    Kostas Papafitsoros, Lukas Adam, Vojtech Cermak, Lukas Picek, (2022) [arXiv preprint]
  • First-order conditions for the optimal control of learning-informed nonsmooth PDEs,
    Guozhi Dong, Michael Hintermüller, Kostas Papafitsoros, Kathrin Völkner, (2022) [arXiv preprint]
  • A social media-based framework for quantifying temporal changes to wildlife viewing intensity: Case study of sea turtles before and during COVID-19,
    Kostas Papafitsoros, Lukas Adam, Gail Schofield, Ecological Modelling, (2023) [DOI
  • Dualization and automatic distributed parameter selection of total generalized variation via bilevel optimization,
    Michael Hintermüller, Kostas Papafitsoros, Carlos N. Rautenberg, Hongpeng Sun, Numerical Functional Analysis and Optimization, 43(8), 887-932, (2022) [DOI]
  • Bilevel training schemes in imaging for total-variation-type functionals with convex integrands,
    Valerio Pagliari, Kostas Papafitsoros, Bogdan Raita, Andreas Vikelis, SIAM Journal on Imaging Sciences, 15(4), (2022) [DOI
  • Optimization with learning-informed differential equation constraints and its applications,
    Guozhi Dong, Michael Hintermüller, Kostas Papafitsoros, ESAIM: Control, Optimisation and Calculus of Variations, 28(3), (2022) [DOI]
  • More aggressive sea turtles win fights over foraging resources independent of body size and years of presence,
    Gail Schofield, Kostas Papafitsoros, Chloe Chapman, Akanksha Shah, Lucy Westover, Liam C.D. Dickson, Kostas A. Katselidis, Animal Behaviour, 190, 209-219, (2022) [DOI]

Public Engagement

With the help of a 2023 QMUL Impact Acceleration Fund, and together with the NGO's ARCHELON and MEDASSET we have created a unique interactive web-platform Zakynthos Turtles, which actively engages visitors and tour operators with responsible sea turtle conservation on Zakynthos Island, Greece and helps towards an adoption of turtle-friendly attitude in ecotouristic activities. Through this platform, visitors upload images of unique turtles that they have observed for individual photo-identification and receive information about them in real time e.g., including their behaviours, interesting stories, and quantification of the human pressure that their observed turtle is subject to (as simplified dissemination of my research).

Since 2024 the project is also supported by the British Chelonia Group.

A related poster presentation presented at the International Sea Turtle Symposium 2024 can be found here.

Back to top