Spiking Neural Networks for Mental Workload Classification with a Multimodal Approach

Abstract--Accurately assessing mental workload is crucial in cognitive neuroscience, human-computer interaction, and real-time monitoring, as cognitive load fluctuations affect performance and decision-making. While Electroencephalography (EEG)-based machine learning (ML) models can be used to this end, their high computational cost hinders embedded real-time applications. This study compares hardware-compatible SNN models with various traditional ML ones, using an open-source multimodal dataset. Our results show that multimodal integration improves accuracy, with SNN performance comparable to the ML one, demonstrating their potential for real-time implementations of cognitive load detection. These findings position event-based processing as a promising solution for low-latency, energy-efficient workload monitoring, in adaptive closed-loop embedded devices that dynamically regulate cognitive demands. Mental workload (also called cognitive load) refers to the mental effort or resources a person uses to perform a task, distinct from the task's external demands [1]. Mental workload classification plays a vital role in enhancing human-computer interaction, advancing cognitive neuroscience, and enabling real-time physiological monitoring. Understanding and accurately assessing cognitive load is essential for optimizing system responsiveness, improving user experience, and ensuring safety in high-stakes environments [2-5]. Traditional approaches have primarily relied on Machine Learning (ML)-based classifiers trained on brain signal datasets such as Electroencephalography (EEG) ones [1, 3, 6].