Techn. Fakultät Willkommen am Institut für Informatik FAU-Logo

Thomas Kautz M. Sc.

Researcher in the Digital Sports group at the Pattern Recognition Lab of the Friedrich-Alexander-Universität Erlangen-Nürnberg

Filtering of position data in sports
  • In recent years, there has been an increasing demand for reliable and objective evaluation of sport specific data. The measurement and analysis of the trajectories of athletes is one possible approach to gain such insights. It allows the assessment of the physical performance and tactical behavior of athletes. Thus, it can yield helpful feedback for athletes, coaches and referees. Furthermore, spectators can be supplied with additional information about the accomplishments of their idols.

    Local Positioning Systems (LPS) provide a means for the measurement of athletes’ positions and motion trajectories. Although such systems can offer accurate position data under ideal measurement conditions, they often suffer from deficient behavior when used at a real sports venue. Since the LPS, which is used in this project, is based on radar, it is susceptible to multipath errors due to reflections of the signal on boards or stands. Besides, weather conditions influence the accuracy of the system. These issues cause deterioration in the reliability of the acquired raw data, so it cannot be used directly for performance or tactical analysis.

    Due to the lack of accuracy in the measurements, further signal processing is necessary to obtain useful information. In this project we try to overcome this problem by combining the measurements with physical constraints, sport specific motion models, plausibility considerations and probabilistic filtering.



Sensor fusion for activity recognition in Rugby
  • The automatic recognition of Rugby-specific activities can yield various important insights and statistics for athletes trainers and fans. Moreover, it can assist referees in their decisions during competitions. In this project, we aim at leveraging symbiotic effects of combining absolute position measurements with inertial measurements in order to identify different Rugby-specific activities reliably. This also includes activities that involve multiple players at the same time, e.g. tackles and scrums.



    Articles in Conference Proceedings
    Workshop on Large-Scale Sports Analytics (21st ACM SIGKDD Conference on Knowledge Discovery and Data Mining), Sydney, Australia, 10.08.2015, pp. 1-4, 2015 (BiBTeX, Who cited this?)
Multimodal characterization of epileptic seizures
  • The purpose of this project is the analysis of electrodermal activity and motion data with respect to important characteristics for the diagnosis of epilepsy. To ensure generalizability, di erent classes of epileptic seizures are taken in to account. Sensor systems which focus solely on motor activities are only applicable for seizures with typical movement patterns (e.g. generalized tonic-clonic seizures). Due to the fact that there are also seizure types showing no typical movements of the patient, it is not be possible to diagnose all types of epilepsy with a motion based sensor system. Previous research suggests, that epileptic seizures often affect the autonomic nervous system. Since changes in the electro-dermal activity  (EDA) are an indicator for the activity of the autonomous nervous system, the measurement and analysis of the EDA signal during epileptic seizures could yield important information about the clinical picture of epilepsy. The aim of the project is to discover characteristics which allows improved detection and characterization of epileptic seizures both in clinical and ambulatory environments. This as a joint project between the Digital Sports Group and the Epilepsy Center at the Erlangen University Hospital.

    Articles in Conference Proceedings
    Heldberg, Beeke E.; Kautz, Thomas; Leutheuser, Heike; Hopfengärtner, Rüdiger; Kapser, Burkhard S.; Eskofier, Björn
    Proceedings of the 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society), Milan, Italy, August, 25-29, pp. 5593-5596, 2015 (BiBTeX, Who cited this?)