Roman Voeikov
Machine Learning Team Lead at Sportradar
London, England, United Kingdom
Summary
👤
Senior🎓
Top SchoolRoman Voeikov is a Machine Learning Team Lead based in London with 8+ years of deep learning experience, specializing in computer vision for sports and real-time systems. He currently leads ML efforts at Sportradar and has driven metric learning for player and team re-identification at Hawk-Eye, combining research-grade publications (CVPR workshop) with production deployment of high-frame-rate, containerized vision pipelines. Roman has a strong experimental mindset—designing multi-task neural networks, optimizing for real-time performance, and integrating filters like Kalman for robust predictions—while also managing cross-team engineering and release workflows. His background in materials science and early lab research gives him a rigorous, data-driven approach to problem formulation and experimentation that complements his open-source work under Oniro.
8 years of coding experience
10 years of employment as a software developer
Machine Learning and Data Analysis, Machine Learning and Data Analysis at CourseraMachine Learning and Data Analysis, Machine Learning and Data Analysis at edXBachelor's degree, Materials Science, Bachelor's degree, Materials Science at Moscow State UniversityMachine Learning and Data Science, Machine Learning and Data Science at Open Data Science
English
Github Skills (18)
depth-estimation10
loss-functions10
depth10
notebook9
makefile9
semantic-segmentation9
segmentation9
docker8
computer-vision8
keras-tensorflow8
keras7
deep-learning7
loader6
neural-network6
tensorflow6
Programming languages (2)
Github contributions (5)
OniroAI/MonoDepth-PyTorch
May 2018 - Sep 2018
Unofficial implementation of Unsupervised Monocular Depth Estimation neural network MonoDepth in PyTorch
Contributions:33 commits, 1 PR, 28 pushes in 3 months
pytorchmonocular-depth-estimationdeep-learningunsuperviseddepth
OniroAI/Semantic-segmentation-with-MobileNetV3
Nov 2019 - Aug 2022
TensorFlow (Keras) implementation of MobileNetV3 and its segmentation head
Contributions:9 commits, 2 PRs, 7 pushes in 2 years 8 months
semantic-segmentationheadkeras-tensorflowperson-segmentationtensorflow
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.