Skills and Qualifications
Natural Language Processing for a Book Publisher
Eksmo is one of the largest publishing houses in Eastern Europe and it needed a natural processing language project....
Neural Network Library
Our machine learning dev team developed a library for neural network construction and compiler optimization. These are...
Face, Sex, Age, Video Emotion Recognition System For NEC
Our machine learning development team has developed various video analytics tools to accurately identify people based...
High-Speed Vehicle Identification System
Our machine learning dev team developed a complex of neural networks that solved the search and recognition problem...
HitFactor is a shooting app that lets users analyze their shooting technique to understand how to improve it. This is...
This was a project for a carsharing service. The project involved creating a system that allowed it to detect and recognize all the faces in the car as well as to detect whether any persons are smoking. The system was required to work offline which was an added challenge. Daniil created a facial recognition solution that could then be embedded into the car’s computer device. He also implemented a cigarette detection solution and a microservice for preparing face vectors to allow for comparison on a device.
Recognition of goods on the shelves
A solution was needed for merchandisers who compare the real arrangement of goods (realograms) with a plan (planogram). OCR and one-shot learning were used to search the product database using triplet loss network training. The generation of images in Blender was required, as was the development of tools marking real goods. Daniil experimented with various detection and classification approaches, taught OCR models, developed image generation using Blender, and ported and optimized models for Android devices.
A project that was required to find athletic starting signals or shots on a soundtrack. Daniil helped develop an architecture using convolutional and recurrent neural networks. He was also required to allow the system to be able to separate shots coming from other directions, which are not important.