Astro Starter Blog

Curriculum Vitae

avatar

Dr. Dmitrij Sitenko

Mathematical Image Processing
Ruprecht Karls University Heidelberg

Education

Ph.D. in Natural Sciences (Magn​a Cum Laude)

Ruprecht Karls Heidelberg University (December 2018 - June 2023)

  • Thesis: Nonlocal Graph-PDEs and Riemannian Gradient Flows for Image Labeling

  • Advisor: Prof. Dr. Christoph Schnörr

  • Specialization: Dynamical formulation of Machine Learning models for Medical Imaging problems on graphs

  • Relevant methods:
    Variational Models, Convex and Nonconvex Optimization, Geometric Methods (Riemannian Manifolds), Deep Learning, Nonlocal PDEs.

  • Overall grade:
    1.0 (Magma Cum Laude)

Master of Science

Karlsruhe Institute of Technology (September 2014- August 2018)

  • Master thesis:
    Time Domain Boundary Integral Equations and Application of the Convolution Quadrature

  • Advisor: Dr. Tilo Arens

  • Specialization:
    Weak formulations of acoustic transmission problems using causal tempered distributions and their numerical solution using discretization of the Laplace domain

  • Relevant methods:
    PDEs, Finite Elements, Inverse Problems

  • Overall grade:
    1.8

Exchange Semester

National Taiwan University, Taipei (August 2016 - February 2017)

  • Specialization: Computational Physics for modeling entangled quantum systems via implementation of Monte Carlo Simulations (C/C++).

  • Major focus on physics

Bachelor’s Degree

Karlsruhe Institute of Technology (September 2011 - September 2014)

  • Bachelor thesis:
    Geometric Rigidity

High School Diploma

Friedrich-List-Gymnasium, Karlsruhe September 2009 - July 2011

Secondary School

Sophie-Scholl-Realschule, Karlsruhe September 2005 - July 2009

Research Interests

My research interests span various areas within applied mathematics and computer science. I am particularly fascinated by:

  • Mathematical Analysis of Machine Learning:

    • Supervised and unsupervised data classification using neural networks.
  • Image Processing:

    • Image denoising and segmentation
      through PDE-based models.
  • Deep Learning:

    • Generative modeling using flow based models.
    • Feature extraction for image labeling and
      classification tasks.
A four node Tree Graph
  • Geometry in Optimization:

    • Application of Riemannian Geometry and Information Geometry to optimization problems on manifolds.
  • Probabilistic Graphical Models:

    • Investigating the application of graphical models for probabilistic reasoning.
  • Optimization Techniques:

    • Convex and combinatorial optimization methods.
  • Parallel Programming:

    • Developing and optimizing algorithms for parallel computing.
  • Nonconvex Programming:

    • Acceleration of optimization algorithms through nonconvex programming.

Publications

  1. A Nonlocal Graph-PDE and Higher-Order Geometric Integration for Image Labeling
    D. Sitenko, B. Boll, C. Schnörr, SIAM Journal of Imaging Sciences, 16(1): 501-567, 2023.
    Link to Paper on Semantic Scholar

  2. Assignment Flow For Order-Constrained OCT Segmentation
    D. Sitenko, B. Boll and , C. Schnörr. International Journal of Computer Vision, 129: 3088-3118, 2021 Link to Paper on Semantic Scholar

  3. Assignment Flows and Nonlocal PDEs on Graphs
    D. Sitenko, B. Boll and , C. Schnörr. In DAGM GCPR, Springer, LNCS , 2021.
    Link to Paper on Semantic Scholar

  4. Assignment Flow for Order-Constrained OCT Segmentation
    D. Sitenko, B. Boll and , C. Schnörr. In DAGM GCPR, Springer, LNCS, 2020
    Link to Paper on Semantic Scholar

Computer skills

  • CUDA
  • C, C++
  • BASH
  • Python
  • Matlab
  • Cython
  • SQL
A four node Tree Graph
A four node Tree Graph

Invited Talks

  1. Assignment Flows and Nonlocal PDEs on Graphs
    In DAGM Germany Conference of Pattern Recognition, 2021

  2. Assignment Flow for Order-Constrained OCT Segmentation
    In DAGM Germany Conference of Pattern Recognition, 2020

  3. Preserving the Geometric Order of Retina Segmentation with Assignment Flow
    In Heidelberg Collaboratory of Image Processing, 2019

Hobbies and Personal Interests

  • Photography: Capturing moments, landscapes, and experimenting with composition and light.
  • Swimming & Jogging: Staying active and maintaining a healthy lifestyle.
  • Cooking: Exploring different cuisines and experimenting with new recipes.
  • Music: Enjoying classical music, playing guitar and piano, and fortunate to share this passion with my wife, a professional musician, who has helped me discover and deepen my love for music.
  • Tech Enthusiast: Passionate about exploring and testing newly released software and applications, as well as implementing the latest research papers to experiment with cutting-edge ideas.
  • Mathematical Art: Exploring the art of problem-solving through mathematical concepts and transforming them into simple, intuitive, and visually engaging art using animations.