E-SPORTS

Explainable models for predicting players’ behaviour and cognitive and emotional state in e-sports.

The E-SPORTS project develops an innovative platform for training e-sports players in the game League of Legends, based on machine learning models capable of analysing both user behaviour and cognitive and emotional state.

The platform integrates interaction data from peripherals and computer vision with psychophysiological measures such as heart rate, electrodermal activity (variation in the skin’s electrical conductivity) and the recording of brain electrical activity through electroencephalography (EEG). This combination of data enables the generation of personalised, adaptive and sustainable training experiences over time.

The project addresses the lack of specific tools focused on well-being in intensive training environments by combining biometrics and artificial intelligence to monitor stress and optimise performance. The solution adapts training in real time, prioritising player well-being while ensuring the reliability, transparency and explainability of the algorithms. Although focused on e-sports, its results are transferable to other existing digital environments such as e-learning.

E-SPORTS, coordinated by Eurecat, involves the participation of the company United Gamers, whose objective is to transform the training of e-sports players through the use of advanced artificial intelligence and data-driven performance optimisation techniques.

The Big Data & Data Science technological unit designs and implements the procedure for collecting and processing physiological, visual and interaction data, and leads the development of explainable models that enables training to be adapted to the user’s behaviour and state.

Meanwhile, the Audiovisual Technologies unit is in charge of the extraction of visual metrics related to facial activity and body movement and collaborates in the development of training adaptation models.