Welcome to the MM4SPA Project

This is the official project page of the MM4SPA joint research project. Check out the project description, project partners (DSHS, L3S, KINEXON) and latest news. To stay updated you can also follow us on Twitter.

Click here for all information on our upcoming Workshop on Multimodal Analysis of Spatio-Temporal Data!




“floodlight” – a high-level, data-driven sports analytics framework

Open source python package for streamlined analysis of sports data released

Post thumbnail
Post thumbnail
Match and performance data from all kinds of sports have seen a remarkable upsurge over the past decade. Better technology and professionalization have led clubs across countries and leagues to collect massive amounts of data and monitor almost every performance related aspect within their organization. However, working with sports data... [Read More]

A Unified Taxonomy and Multimodal Dataset for Events in Invasion Games

A hierarchical taxonomy that enables a minimal and objective annotation. And is modular expendable to fit the needs of various invasion games

Post thumbnail
Post thumbnail
Events like passes or shots play an important role for the interpretation of complex invasion games like soccer, handball, hockey, or basketball. Over the last years, three fundamental perspectives emerged with regard to the analysis of sports games, which all value different characteristics of the respective sports. Somewhat overstated: The... [Read More]