The L3S Research Center and the Institute for Training Science and Sports Informatics at the German Sport University Cologne will analyze sports data with artificial intelligence. The BMBF is providing the MM4Sports project with more than 800,000 euros over a two-year project period (2020-2022) as part of the funding line "Application of artificial intelligence methods in practice."
Another project partner is the Munich-based technology company KINEXON, which specializes in the real-time collection and processing of sports performance data. The Institute for Training Science and Sports Informatics has been cooperating with KINEXON for two years. Therefore, it is one of the few research facilities worldwide where position data in indoor and outdoor sports can be collected. "Based on the existing position data solution of KINEXON, an automated consolidation of position and video data in sports games is missing so far, which is important for basic and application-oriented research projects in the future market of sports analytics," says Univ.-Prof. Dr. Daniel Memmert, managing director of the Institute for Training Science and Sports Informatics at the German Sport University Cologne. He and his team have been working for years on the analysis and simulation of complex position data and have developed and validated theoretical framework models for this purpose
The aim of the new project is now to integrate machine learning methods into the existing analysis platform of KINEXON. "The core of our solutions is our in-house sensor network for the acquisition of precise and reliable live data in indoor and outdoor environments. In the future, we want to push more strongly the integration of additional data sources in real-time and further develop the automatic translation of data into well-founded analyses and recommendations for action", says Dr. Maximilian Schmidt, Managing Director of KINEXON Sports & Media.
The L3S provides expertise in handling video data and information in the age of internet and digitalisation. Prof. Dr. Ralph Ewerth and his "Visual Analytics" research group have been working for many years on the analysis of videos and multimodal data. Among other things, the group researches and develops methods of automatic video analysis and artificial vision in combination with machine learning processes. "The automatic enrichment and visualization of position and video data, as well as their analysis by machine learning methods, supports sports analysts in the preparation of important sport-relevant questions regarding team and individual tactics. This can increase the success of teams in high-performance sports," says Prof. Ewerth, outlining the area of application. The project initially uses data from soccer, basketball, and handball.