The ability to predict the future movement of an agent has many applications, from anticipating behaviours within an autonomous driving environment to detecting abnormal behaviours in a security setting. Within sports, being able to predict paths has applications for match analysis, strategy planning, and automated broadcasting.
Path prediction is typically performed at an individual level, predicting the future path of each person (or the ball) in turn. While such predictions may consider the locations and past movements of other players (both teammates and opponents), as movements are coordinated across a team information is lost by predicting paths in this manner. This project will explore how these existing path prediction approaches can be extended to predict the paths of multiple people (and potentially the ball) jointly.
As part of the research project, you will:
- investigate and deploy machine learning algorithms to predict the future motion of multiple players (and potentially the ball) simultaneously in a sporting environment
- investigate improvements to the deployed algorithms to enhance performance
- benchmark developed algorithms and measure accuracy/performance.
Exact task details will be determined in consultation with the project supervisors.
The outcomes of this work will include:
- one or more algorithms that are trained and evaluated on sports trajectory data
- performance benchmarks for the deployed algorithm(s).
Skills and experience
Strong programming experience, especially with Python, is ideal.
Some prior machine learning or computer vision experience is desirable, but not mandatory.
You may be eligible to apply for a research scholarship.
Contact Dr Simon Denman for more information.