Vladimir Loginov
Computer Vision Engineer at Tinkoff
Saint Petersburg, Saint Petersburg, Russia
Summary
🤩
Rockstar🎓
Top SchoolVladimir Loginov is a computer vision engineer with 8 years of experience building and shipping deep learning solutions across automotive ADAS and model deployment at scale. He developed a state-of-the-art scene text recognition network and contributed formula-recognition demos to Intel’s Open Model Zoo, including an im2latex demo and datasets for handwritten math. At АТОМ he designed a camera-based parking-slot detection system fused with ultrasonic sensors and maintained in-vehicle pipelines using Docker and Ansible, and now works on image generation at Tinkoff. Comfortable bridging research and production, he has a track record of squeezing extra performance from models (up to ~2% gains) and adapting algorithms to real-world constraints like fisheye optics and top-down detection. Off-duty he’s a corgi owner and a devoted VSCode enthusiast.
8 years of coding experience
7 years of employment as a software developer
Bachelor's degree, Applied mathematics and Informatics, Bachelor's degree, Applied mathematics and Informatics at State University of Nizhni Novgorod named after N.I. Lobachevsky (UNN)Master's degree, Big Data and Machine Learning, Master's degree, Big Data and Machine Learning at ITMO University
English, Spanish, Russian
Github Skills (7)
inference10
deep-learning10
python10
openvino10
modello9
cnn-model9
computer-vision9
Programming languages (7)
Github contributions (5)
openvinotoolkit/open_model_zoo
Jul 2020 - Jan 2022
Pre-trained Deep Learning models and demos (high quality and extremely fast)
Role in this project:
ML Engineer
ML EngineerContributions:114 reviews, 266 commits, 21 PRs in 1 year 6 months
Contributions summary:Vladimir implemented a demo for mathematical formula recognition, specifically focusing on the im2latex task. Their work involved adding and modifying files related to the demo, including core files like `im2latex_demo.py` and `utils.py`, and making changes for image preprocessing and model inference. They also updated the code to support model input and output layers, enabling the functionality of the model.
deep-learningmodelscaffemodeltensorflow-modelsmodel-zoo
morkovka1337/open_model_zoo
Jul 2020 - Jan 2022
Pre-trained Deep Learning models and samples (high quality and extremely fast)
Contributions:2 reviews, 2 PRs, 342 pushes in 1 year 6 months
pytorchdeep-learningpre-trainedhigh-qualitydeep-learning-models
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.