We attended BOSA2024! What? Why? Who?
We attended BOSA!… which is a week-long course for data scientists organized by the IDLab Sports Data Science teams of Ghent University and Antwerp University. As a result, we are richer with new skills in the use of artificial intelligence (XAI) explanatory methods, knowledge of machine vision techniques, applications of biomechanics principles and other practical applications of AI in sports. NASK SCIENCE was represented by Michal Kozbial.
Data analysis in sports is definitely changing the game. Data analysis helps sports entities evaluate the performance of their athletes and assess the recruitment necessary to improve the team’s performance. It also evaluates the strong and weak areas of their opponents, enabling coaches to make the right decision on their tactics. But data is not only used in performance optimization. It is also used in the storytelling of the game and the safety monitoring of the athlete/venue.
The Becoming Outstanding In Sports Analytics (BOSA) Winter School, was organized by the IDLab Sports Data Science teams of Ghent University and the University of Antwerp. It was a week-long intensive course bringing together junior and senior academics and practitioners in the field of sports data science. Two tracks were organized in parallel:
- a track focused on sports scientists that want to gather more AI/data science skills
- a more hands-on track for computer scientists to apply their expertise/skills in a sports context.
The Winter School was open to people coming from academia and professionals active within the sports sector.
During the first three days, BOSA participants gained fundamental and practical skills from both disciplines (sport and data science) and combined/mixed them in the final hackathon project in which they had intensively worked together in a multidisciplinary context. The hackathon required participants to put together the diverse skill sets they bring to the table. During the hackathon, they were in close contact with the end users/federations that define the hackathon topics.
Michał Koźbiał’s, who represented NASK during the event, research interests include silhouette-based age estimation, baseline sample detection in biometrics, and feature point detection on the human body. His research interests include image processing, silhouette biometrics, deep learning networks, and generation networks. He is a doctoral student at the TIB PAN Doctoral School.