Jan Glinko

Eye tracking with deep learning - from research and POC to final end-to-end pipeline. Created gaze estimation network outperforming SOTA.

Optimizing neural networks performance on Google Coral (TPU)

Conducting research studies and teaching students

I took activate part in 2 computer vision Kaggle competitions:

Stroke Etiology Classification: top 4% solution,

Blood Vessels Segmentation: did not choose solution that would give me top 5%,

and 1 tabular data competition:

Parkinson progression prediction

Ongoing computer vision competition:

Abdominal Trauma Detection

Object detection, semantic segmentation, synthetic datasets generation, stereo vision, path planning, robotics, ROS

I wrote two scientific papers:

https://link.springer.com/chapter/10.1007/978-3-031-16159-9_12 (4 citations)

https://link.springer.com/chapter/10.1007/978-3-031-16159-9_13

Face detection and anonymization on large scale aerial photos.