This project focusses on development of sports related AI-driven video analytics. Research contributions will focus on:
- the development of convolutional neural networks to detect in-game events (goals, points, throw-ins, penalties, etc);
- detection of individual player behaviours (goal scorers, player infringements, player fitness, etc), building on video-based behavioural analysis research conducted in the ISRC;
- generation of realistic defensive and attacking plays based on the ball and team movements using a generative adversarial network to learn spatio-temporal interactions between players' movements and subsequently simulate how the opposing team will react to a newly developed strategy.
The successful candidate will have the opportunity to work at the cutting edge of video sports analytics alongside a well-established CCTV company resulting in high quality publications.