Multimodale Analyse für die Sportanalytik

Intelligente Synchronisation und semantische Anreicherung von Positions- und Videodaten für die Analyse von Sportspieldaten

The MM4SPA project represents a major step forward in the field of sports analytics by integrating advanced machine learning methods into the existing analytics platform of industry partner KINEXON. This innovative project, led by the Institute for Exercise Training and Sports Informatics in collaboration with the Visual Analytics Research Group, aims to develop new, innovative services in the field of sports analytics.

By integrating sensor systems and video data, the MM4SPA project is able to semantically enrich data through automatic detection of sport-specific events and their evaluation using collaboratively developed expert models. This approach enables the provision of digital assistants to support customer decision making in the field of sports analytics.

The project focuses on the disciplines of football, basketball and handball and aims to develop a flexible and future-proof system capable of adapting to the varying complexity and dimensionality of the data associated with each sport. By using video analysis, computer vision and machine learning methods, the MM4SPA project will provide valuable insights to sports analysts and increase the success of teams in high-performance sports.

In addition to the technical advancements, the MM4SPA project will also promote knowledge transfer between industry and academia, and ensure that the solution can be independently developed by small and medium-sized companies in the future. The project represents a significant milestone in the field of automated match analysis and data analysis in sport, and has great potential to advance our understanding of sports performance.


Prof. Dr. Ralph Ewerth

Head of Visual Analytics Research Group