Who Is Playing? An Innovative Continuous Identification & Monitoring System in the Esports Industry

In recent years, esports has experienced exponential growth, becoming one of the most promising sectors of the entertainment industry. With millions of fans around the world and prize money that rivals that of traditional sporting competitions, esports has gained a prominent place in contemporary popular culture. However, this growing popularity has brought with it new challenges. Among these, the need to ensure a reliable system of player identification in online competitions. With increasing numbers of participants and the relevant presence of misconduct, it is essential to implement advanced systems to ensure the integrity and equity of competitions. In response to this need, this thesis, in collaboration with Spike Reply, proposes the development of an innovative system designed to ensure continuous identification and monitoring of players during competitions. This is achieved through a modular architecture that integrates several essential components. The first critical element of the system is a face detection module, which uses state-of-the-art algorithms to identify the faces of players within the game environment in a timely and accurate manner, with negligible impacts on performance. This component provides a critical foundation for continuous identification, ensuring that each player is correctly identified throughout the game. Once the players' faces have been detected, the system uses the biometric characteristics to determine the identity of the participant. This biometric recognition capability allows the system to reliably verify that a person is who has registered for the competition. The system also includes a liveness detection module whose main role is to protect the identification process from possible spoofing or fraud attempts, by understanding that what is captured by the webcam is real or fake. This feature is crucial for maintaining trust on the identification process, ensuring that only real and legitimate users are granted access. Finally, the system is complemented by the event manager feature, which play a crucial role in the continuous monitoring process. This module can notify any event that may indicate potential misconducts, allowing for timely checks and optimizing human assessment activities. Its ability to respond quickly to potentially fraudulent situations help ensuring that only legitimate participants can compete to an online tournament. In conclusion, the primary outcome of this thesis is the development of a reliable system aimed at addressing the identified needs of continuous player identification in esports. While the system provides a solid foundation, there is also room for evolution and improvement. This work offers an in-depth exploration of the theoretical considerations and practical solutions, providing insights into the complexities of continuous identification systems in esports.