Yaroslav Prytula

I am interested in solid theoretical and algorithmic foundations for learning and computational decision-making under imperfect information. To achieve that, I combine techniques from machine learning, optimization, and statistics.

My research focuses on instance-aware segmentation, transformer-based models, and biomedical imaging, with an emphasis on robustness and overlapping objects.


Some current research interests

Medical Computer Vision

Query-based and transformer-driven segmentation models for overlapping objects in biomedical images.

Publications on this topic:

IAUNet architecture

IAUNet: Instance-Aware U-Net

Yaroslav Prytula, Illia Tsiporenko, Ali Zeynalli, Dmytro Fishman

In CVPR Workshops 2025