SANTICCHI GABRIELE

Deep Learning and Sensor Fusion for Automated Sports Performance Analysis - Application of deep learning and sensor fusion techniques for automated technical performance analysis in football. This research involves the development of neural network architectures (CNN-LSTM) for automatic detection and classification of technical events (kicks, passes, ball controls) from inertial sensor signals, the integration of geofencing algorithms for spatial event contextualization, and the implementation of Bayesian fusion approaches combining IMU and GNSS data to enhance tactical classification. It also includes regression models for ball speed estimation and automated passing network analysis. (2022 - 2024)

Wearable Sensor Validation for External Load Monitoring in Football - Validation of wearable technologies (GNSS and IMU) for external load monitoring in football. This research focuses on assessing accuracy and inter-device agreement among commercial tracking systems for the quantification of physical performance metrics (distance covered, speed zones, accelerations/decelerations) under ecologically valid training and match conditions. It includes the development and validation of step detection and distance estimation algorithms using shank-mounted inertial sensors, enabling continuous monitoring even in GPS-denied environments. (2024 - 2026)