Accessible Video Analytics: the Use Case of Basketball

The development of detailed and accurate analytics is nowadays considered among the essential elements that drive the preparation of a competition. In particular, video analytics contribute by providing both the athletes and the technical team an all-round perspective on the opponent through multiple recordings, in different environmental conditions and configurations, observing the technical and tactical performances across time. However, the chance of collecting video analytics is generally linked to the availability of suitable hardware and software, to collect, store, and analyze the video streams, so as to extract the relevant information. This is often considered as a privilege reserved to professional teams, which can also benefit from the presence of a dedicated personnel assigned to this task. In this paper we present a study that demonstrates that the recent developments in computer vision and machine learning have enabled the possibility of deploying accessible solutions, yet retrieving reasonable and interpretable analytics.