Kyriakos Flouris

Decoding the Universe with Code, Curved Spaces, and Quantum Ideas

Welcome to my digital portfolio! I am a Group Leader for Biomedical Machine Learning at the Biostatistics Unit, University of Cambridge , where I merge the realms of Artificial Intelligence, biostitistics, medical imaging, and theoretical physics. From developing novel generative models to exploring the mysteries of quantum physics and fluid dynamics, my work is driven by a fascination with the intersection of technology and nature. If you enjoy thinking about how AI can shape our understanding of the world, you are in the right place.

Research Interests

• Generative Modeling and AI (Gradient flows, Normalizing Flows, Variational Autoencoders, Diffusion Models)
• Medical Imaging (Computed Tomography, Radiomics, MRI Super-Resolution)
• Differential Geometry and Theoretical Physics
• Fluid Structure Interaction and Lattice Boltzmann Methods
• Quantum Machine Learning

Education

• Doctor of Sciences, ETH Zurich (2015-2019)
  • Thesis: Flow and Dirac Particles on Curved and Topological Manifolds
  • Developed lattice Boltzmann and Dirac equation solvers for fluid-structure interactions and quantum systems
• Master of Science, University of Cambridge (2013-2014)
  • Part III of Physical NatSci Tripos - First Class (70.0% GPA)
  • Courses in Particle Physics, Relativistic Cosmology, Gauge Field Theory, and Non-Linear Optics
• Bachelor of Science, University of Cambridge (2010-2013)
  • Experimental and Theoretical Physics (2.1 - 67.5% GPA)

Professional Experience

• Senior Research Scientist, Computer Vision Lab, ETH Zurich (2019-present)
  • Developing and applying deep learning techniques in medical imaging
  • Leading projects on generative modeling, CT radiomics, and MRI super-resolution
  • Exploring quantum-inspired machine learning algorithms and neural PDE solvers
  • Supervising MSc and PhD research projects, and teaching at summer schools
• Research Assistant, Computational Physics for Engineering Materials, ETH Zurich (2015-2019)
  • Developed fluid structure interaction models with lattice Boltzmann methods
  • Researched Dirac fermions and topological matter
  • Teaching assistant in Python programming and computational methods
• Research Assistant, Institute of Astronomy, ETH Zurich (2014-2015)
  • Part of the MUSE project at the Very Large Telescope (VLT)
  • Developed algorithms for data reduction and analysis

Selected Publications

• "Canonical Normalizing Flows for Manifold Learning", NeurIPS, 2023
• "Explicitly Minimizing the Blur Error of Variational Autoencoders", ICLR, 2023
• "Curvature-Induced Quantum Spin-Hall Effect on a Möbius Strip", Physical Review B, 2022
• "Assessing Radiomics Feature Stability with Simulated CT Acquisitions", Scientific Reports, 2022
• "A Matrix Product State Model for Simultaneous Classification and Generation", Quantum Machine Intelligence, 2024

Skills

• Programming Languages: Python, C++, MATLAB, Mathematica, LaTeX, Unix, GitHub
• Tools and Frameworks: PyTorch, Deep Learning, Generative Modeling, Lattice Boltzmann Methods
• Research Skills: Data Acquisition and Analysis, Theoretical Research, Teaching

Awards and Honors

• Master of Science, First Class Grade, University of Cambridge
• PHRT Funding, ETH Zurich - Radiomics and DeepFlow Projects
• Highest International Mark in Physics (GCE A-Level), Edexcel

Personal Interests

• AI and Numerics
• Theoretical Physics
• Skiing and Mountain Biking
• Design and Technology
• Football and Running

Thank you for visiting!

Kyriakos Flouris