The goal is to develop cognitive training for sports using Machine Learning. Using data extracted from frames of video of soccer matches, the project is to see whether ML can be used to determine patterns of player actions that lead to a goal. These patterns would be used in a virtual setting to train young soccer players to quickly understand the situations and follow the pattern resulting in a goal


Data is extracted from a soccer match video that shows for every video frame the location on the field of each player and referee. It also shows the ball location in three-dimensional space. The idea is to learn offense patterns used before a goal is scored that will be used in a virtual reality simulation to be used by soccer coaches. This is still being worked but looking at Long Short Term Memory in Deep Multi-agent Imitation Learning.


  • Sports

Tools Used

  • Data Science Experience (DSX)

Data Science Techniques

Long Short Term Memory in Deep Multi-agent Imitation Learning


  • Ability to use data from soccer matches to detect player patterns of the offensive soccer players leading to goals.  This data will be used for virtual reality training.

Need help with a similar case?

Contact us to find out how we can help!