A Novel Approach for Automatic Detection of Driver Fatigue Using EEG Signals Based on Graph Convolutional Networks
DOI: 10.3390/s24020364
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This study addresses the critical safety issue of driver fatigue, a major contributor to traffic accidents, by proposing a novel multi-class automatic detection system. Previous research largely relied on binary classification (fatigued vs. normal), manual feature extraction, and self-report questionnaires, which limit real-world operational capability and accuracy. To overcome these limitations, the authors developed an end-to-end deep learning framework using electroencephalography (EEG) signals to detect five distinct levels of passive work-related fatigue. The system aims to provide more granular fatigue assessment and higher reliability by incorporating physiological confirmation beyond subjective reports. The methodology involved designing a cost-effective driving simulator using Logitech hardware and a 50-inch screen to replicate highway driving conditions without traffic, thereby inducing mental fatigue. To ensure ecological validity, the simulator included stereo speakers playing recorded engine and cabin noises at realistic volumes. Data were collected from 20 participants (10 men, 10 women, aged 20–35) who underwent fatigue-inducing driving sessions. Physiological signals, including EEG, ECG, electromyography (EMG), and respiratory effort, were recorded. Fatigue stages were validated using these physiological changes rather than solely relying on self-report questionnaires. The core technical approach combines Generative Adversarial Networks (GAN) for data augmentation to address limited dataset sizes and Graph Convolutional Networks (GCN) for feature extraction and classification. The proposed GCN architecture consists of five convolutional graph layers, one dense layer, and one fully connected layer, designed to handle the graph-structured nature of EEG data and resist environmental noise. The results demonstrate high performance across four practical application cases, with classification accuracies of 99%, 97%, 96%, and 91%. The model successfully distinguished between five fatigue classes, outperforming previous methods that typically achieved below 90% accuracy for multi-class tasks or relied on binary classification. The integration of GANs effectively augmented the dataset, while the GCN provided robust feature learning without the computational complexity associated with high-channel-count EEG setups. The study confirms that physiological indicators, particularly EEG patterns, change predictably before the onset of severe sleepiness, allowing for early detection. The significance of this work lies in its transition from binary to multi-class fatigue detection, offering a more nuanced understanding of driver weariness that is better suited for real-world implementation. By eliminating reliance on self-reporting and reducing computational load through efficient deep learning architectures, the proposed system offers a viable solution for real-time driver monitoring. The findings suggest that combining GANs for data enhancement with GCNs for spatial-temporal feature extraction provides a superior approach to EEG-based fatigue detection, potentially reducing accident rates through timely, accurate alerts.
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-08-09 |
| archive | success | openalex | — | — | 5 | 2026-08-09 |
| extract | success | pdftotext | — | — | 4 | 2026-08-10 |
| clean | success | clean | — | — | 2 | 2026-08-10 |
| chunk | success | chunk | — | — | 2 | 2026-08-10 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 2 | 2026-08-10 |
| promote | success | — | — | — | 1 | 2026-08-09 |
| summarize | success | llm | qwen3.6-27b-nvidia | summ-v5 | 2 | 2026-08-10 |
| tag | success | vector_similarity | — | — | 17 | 2026-08-11 |
| verify | success | — | — | — | 2 | 2026-08-10 |
Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Empirical Findings: physiological data
- Methodological Resource: tool software