I am a biomedical imaging scientist and machine learning engineer who has
8 years of experience in analyzing high throughput biomedical imaging and multi-omics data using deep neural networks, machine learning, parallel programming, and
2 years of experience in enterprise software development, server side programming, relational databases.
Server side programming, databases: SQL, SQL Server, Java EE, Spring MVC, Object Relational Mapping.
Version control systems: Git, SVN.
Operating systems: Linux (debian derivatives)
Writing research papers, giving presentations, efficient collaboration and communication.
Publications
I co-authored more than 15 publications while working as a biomedical imaging scientist.
See my Google Scholar page for the details.
Employment and experience
Machine learning engineer, applied scientist (deep neural networks for microscopy image analysis), part time Single-Cell Technologies Ltd. 2022-
Efficient implementation of deep neural network based computer vision algorithms for instance segmentation and classification of microscopy and pathology images in 2D and 3D with Tensorflow C API and C++, OpenCV, CUDA.
Image analysis pipelines (instance segmentation, classification, tracking) for high-throughput microscopy images of single-cells in 2D and 3D for several systems biology research projects.
Building deep neural networks & training object recognition pipelines for microscopy image analysis from scratch (Tensorflow, PyTorch)
Ultra-fast deep learning based single-cell segmentnation and tracking for 3D multiphoton microscopy images.
Efficient implementation of a differential intensity contrast microscopy image reconstruction algorithm with OpenCL.
Development of synthetic data generation pipelines with image style transfer and GANs for 2D instance segmentation in human tissues.
Development of a novel 3D active contour method for nuclei segmentation in 3D, numerical approximation with finite differences, efficient implementation in CUDA and C++, 3D visualization with MITK, Qt, visualization with vispy.
Worked on the statistical analysis of multi-omics data (genomics, proteomics) in several projects, used several machine learning techniques for feature selection (in a large population health database) such as Lasso and ElasticNet.
Collaboration in several international research consortiums.
Analysis of social networks using graph algorithms and machine learning: personalized PageRank, power iteration, random walks, logistic regression, sentiment analysis.
Analysis of social media posts using conditional random fields, discriminative modeling, feature extraction, n-grams, L-BFGS.
Development of content management systems with 3-tier architecture in Java Enterprise Edition (EJB 3.0)
Planning and implementation of entity relationship mappings, development of database schemas in SQL Server, schema normalization, refactoring T-SQL stored procedures, caching, load balancing, http proxy servers.
Development of web interface with sping MVC, HTML, CSS and jQuery.
Education
Doctor of philosophy (Ph.D.) in Computer Science University of Szeged 2017-2021
Supervisor: Dr. Peter Horvath
Thesis: Active contour and deep learning methods for single-cell segmentation in microscopy images (2023)