Modified TSception for Analyzing Driver Drowsiness and Mental Workload from EEG

2025-12-25

AI Summary

To address the real-time monitoring of driver fatigue and cognitive workload, this paper proposes an enhanced TSception deep learning framework to improve robustness and cross-task generalization in EEG signal modeling. Methodologically, it introduces a five-layer temporal refinement architecture, an adaptive average pooling mechanism, and a two-stage spatiotemporal feature fusion strategy to enable multi-scale temporal modeling and joint spatiotemporal optimization. Evaluated on the SEED-VIG dataset, the model achieves 83.46% accuracy (95% CI ±0.24); on the STEW dataset, it attains state-of-the-art (SOTA) performance with 95.93% and 95.35% accuracy for binary- and ternary-class mental workload classification, respectively. This work delivers a high-reliability, end-to-end solution for in-vehicle real-time EEG monitoring systems, demonstrating superior generalization across diverse cognitive states and experimental protocols.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Hardware-aware ML
• Computer Vision: Multi-modal Vision

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

Driver drowsiness remains a primary cause of traffic accidents, necessitating the development of real-time, reliable detection systems to ensure road safety. This study presents a Modified TSception architecture designed for the robust assessment of driver fatigue using Electroencephalography (EEG). The model introduces a novel hierarchical architecture that surpasses the original TSception by implementing a five-layer temporal refinement strategy to capture multi-scale brain dynamics. A key innovation is the use of Adaptive Average Pooling, which provides the structural flexibility to handle varying EEG input dimensions, and a two - stage fusion mechanism that optimizes the integration of spatiotemporal features for improved stability. When evaluated on the SEED-VIG dataset and compared against established methods - including SVM, Transformer, EEGNet, ConvNeXt, LMDA-Net, and the original TSception - the Modified TSception achieves a comparable accuracy of 83.46% (vs. 83.15% for the original). Critically, the proposed model exhibits a substantially reduced confidence interval (0.24 vs. 0.36), signifying a marked improvement in performance stability. Furthermore, the architecture's generalizability is validated on the STEW mental workload dataset, where it achieves state-of-the-art results with 95.93% and 95.35% accuracy for 2-class and 3-class classification, respectively. These improvements in consistency and cross-task generalizability underscore the effectiveness of the proposed modifications for reliable EEG-based monitoring of drowsiness and mental workload.

Problem

• Develops a real-time EEG system to detect driver drowsiness for road safety
• Introduces a modified neural network for stable EEG-based fatigue assessment
• Enhances model generalizability to also evaluate mental workload from EEG

Innovation

• Hierarchical architecture with five-layer temporal refinement strategy
• Adaptive Average Pooling for handling varying EEG input dimensions
• Two-stage fusion mechanism optimizing spatiotemporal feature integration