Artificial Intelligence-Enhanced Brain-Computer Interface Framework for Real-Time User-Responsive Experience

Wrocław University of Science and Technology 2024 ZPI Day Submission49 Authors

Keywords: BCI, Brain-Computer Interface, Unity, AI, Classificators, Machine learning, Unity Framework, EEG data, Neurofeedback, Real-Time Interaction, Game Personalization, Engagement Monitoring

TL;DR: Mentalizer framework enables real-time cognitive state analysis using EEG signals, offering a modular, adaptable system for applications in gaming, workplace productivity, and other experiences by dynamically adjusting based on emotional states.

Abstract:

Brain-computer interface (BCI) technology holds immense promise for real-time adaptive systems. However, current tools often fail to effectively integrate cognitive and emotional responses into practical applications. To address this, we developed $\textit{Mentalizer}$ - modular BCI framework that processes EEG signals and integrates seamlessly with Unity-based applications.
$\textit{Mentalizer}$ includes independent modules for device interfacing, preprocessing, classification, and application integration, all configurable via Unity’s Scriptable Objects. This design ensures flexibility and adaptability to diverse use cases. The framework classifies cognitive states such as boredom, flow, and frustration, enabling dynamic adjustments in applications.
As part of the project, we conducted experiments to gather data to train classifiers and validate the framework in practical scenarios. Use cases include dynamically adjusting video game difficulty to maintain player engagement and detecting attention lapses in high-stakes professions like airport security. $\textit{Mentalizer}$ demonstrates potential to enhance productivity, user satisfaction, and engagement across various domains.

Submission Number: 49