Company: United Developers

We are United Developers, an international IT company with over 10 years of experience in developing and supporting mobile solutions for clients from Europe and the USA. Our key expertise lies in creating complex mobile applications for IoT (Internet of Things) and integrating AI/ML into products for businesses and startups.

Project: Automatic control of a PTZ camera for American football tracking. Computer Vision + C++.

Tasks:

• Training a YOLO model for ball detection (from scratch)
• Hybrid tracking logic: switching between players and the ball
• Adapting the C++ pipeline
• Testing on real videos from PTZ cameras
• Optimization for OpenVINO

Requirements:

• C++ for production code (modern standards, multithreading, memory)
• OpenCV, video processing (capture, frame processing)
• Understanding of object tracking (filters, data association)
• Experience training detection models (YOLO / PyTorch) - important, but less experience is acceptable than in C++

Plus:

• OpenVINO / ONNX Runtime
• Sports analytics / ball tracking

Format: remote, part-time (~20h/week)

Duration: ~2 months (June — August 2026)

Compensation: ~220,000₽ fixed for the project (discussed based on test task results)

Test Task: paid

What's already available:

• Working player detection model
• C++ pipeline (adapted from soccer phase)
• Video dataset from the client
• Ball tracking code from a basketball project.