Eugene Khvedchenia
Senior AI Software Engineer at NVIDIA
Odesa, Odesa Oblast, Ukraine
Summary
🤩
Rockstar🎓
Top SchoolEugene Khvedchenia is a Senior AI Software Engineer with over 15 years of hands-on experience in computer vision and deep learning, now contributing at NVIDIA after roles at Deci AI and multiple CV-focused startups. He’s a Kaggle Grandmaster and prolific open-source maintainer—co-author and core contributor to the widely used albumentations library and author of pytorch-toolbelt—bringing practical tooling improvements like multi-threading and memory-efficient loss functions. His background spans production C++ integrations (OpenCV KAZE/AKAZE), AR and SLAM systems, OCR and large-scale image retrieval, evidencing a rare blend of low-level performance engineering and high-level ML research. Based in Odesa, Ukraine, he excels at solving open-ended problems and translating research-grade algorithms into robust, tested code for real-world systems.
15 years of coding experience
13 years of employment as a software developer
Masters Computer Software Engineering, Masters Computer Software Engineering at Odessa State Academy of Refrigeration
English
Github Skills (19)
unit-testing10
pytorch10
opencv10
c-language10
bounding-box10
pytest10
python10
image-processing10
testing10
machine-learning10
numpy10
cplus10
object-detection10
computer-vision10
image-augmentation10
Programming languages (4)
Github contributions (5)
BloodAxe/pytorch-toolbelt
Mar 2019 - Jan 2023
PyTorch extensions for fast R&D prototyping and Kaggle farming
Role in this project:
ML Engineer
ML EngineerContributions:32 releases, 1 review, 1053 commits in 3 years 11 months
Contributions summary:Eugene's contributions focused on implementing, refining, and optimizing various loss functions within the PyTorch-based machine-learning toolkit. They implemented new loss functions, optimized existing ones for memory efficiency and added support for features like label smoothing. The user also worked on unit tests for implemented classes, and refactoring.
kagglepythonimage-segmentationprototypingextensions
albumentations-team/albumentations
Aug 2018 - Jun 2022
Fast and flexible image augmentation library. Paper about the library: https://www.mdpi.com/2078-2489/11/2/125
Role in this project:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:2 releases, 81 reviews, 113 commits in 3 years 10 months
Contributions summary:Eugene made significant contributions to adding support for image augmentation transformations related to bounding boxes. They integrated image augmentation techniques like IAAAffine, IAAFlipud, IAAFliplr, and IAACropAndPad, for enhanced object detection capabilities within the `albumentations` library. The user focused on providing keypoint transformation functionality, including fixes for the same transformations. They also fixed issues related to shape preservation and implemented multi-threading support for the library.
fast-augmentationspythonalbumentationsdetectiondeep-learning
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.