| KEYWORD |
Artificial Intelligence techniques for analyzing tennis players' performance
Tesi esterna in azienda
Parole chiave DEEP LEARNING, INTELLIGENZA ARTIFICIALE, RETI NEURALI, SPORT, VISIONE ARTIFICIALE
Riferimenti FABRIZIO LAMBERTI
Gruppi di ricerca DAUIN - GR-09 - GRAphics and INtelligent Systems - GRAINS
Tipo tesi TESI IN AZIENDA, TESI IN COLLAB. CON AZIENDA
Descrizione Several new thesis proposals are available, which will be developed in collaboration with HyperTennis (https://hypertennis.it/) a startup / spin-off of Politecnico di Torino. HyperTennis is designing solutions for the application of Artificial Intelligence and other technologies to analyze and improve the performance of tennis players. Thesis works will require the application of Computer Vision and Deep Learning techniques to:
1) perform high-speed, Human Pose Estimation and extract the articulated 3D model of the tennis players from the video feed coming from multiple, synchronized cameras
2) identify relevant tennis players' actions in the execution of technical gestures, i.e. shots (serve, forehand, backhand, etc.)
3) extract accurate information about the ball (trajectories and impact with the racket and the field). 4) automatically classify shot type
5) identify/separate specific phases in the execution of a shot
6) predicting next shot based on historical data
Vedi anche http://grains.polito.it/work.php
Scadenza validita proposta 27/11/2024
PROPONI LA TUA CANDIDATURA