Denis Prokopenko
PhD Candidate
London, England, United Kingdom
Summary
🤩
RockstarDenis Prokopenko is a PhD candidate and experienced software engineer with nine years of hands-on experience in machine learning engineering and research-based development. Based in London, he contributes to open-source ML tooling—most notably improving image-to-image evaluation metrics by migrating Kernel Inception Distance to PyTorch and fixing MS-SSIM issues in the photosynthesis-team/piq repository. His work bridges research rigor and production practicality, ensuring metric implementations are both efficient and numerically reliable. Comfortable navigating NumPy/sklearn-to-PyTorch transitions, he brings deep familiarity with model evaluation pipelines and reproducible code. Colleagues can expect a researcher-engineer who prioritizes correctness and pragmatic engineering trade-offs when turning academic ideas into usable tools.
9 years of coding experience
Github Skills (3)
python10
pytorch10
numpy6
Programming languages (2)
Github contributions (5)
photosynthesis-team/piq
Apr 2020 - Jan 2023
Measures and metrics for image2image tasks. PyTorch.
Role in this project:
ML Engineer
ML EngineerContributions:12 releases, 87 reviews, 42 commits in 2 years 9 months
Contributions summary:Denis primarily focused on enhancing the Kernel Inception Distance (KID) metric within the image2image task framework. Their contributions involved migrating the KID implementation from NumPy and sklearn to PyTorch, including necessary code adjustments. Furthermore, the user addressed and fixed issues related to the Multi-scale Structural Similarity (MS-SSIM) feature, ensuring valid metric values.
vifbrisquemsepsnrimage-quality
denproc/mrirecon
Mar 2022 - Sep 2023
MRI Reconstruction Toolbox PyTorch
Contributions:2 releases, 3 PRs, 15 pushes in 1 year 6 months
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.