Driver Fatigue Classification With Independent Component by Entropy Rate Bound Minimization Analysis in an EEG-Based System
DOI: 10.1109/jbhi.2016.2532354
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 leading cause of motor vehicle accidents, by developing an automated electroencephalography (EEG)-based classification system. The research aims to improve the accuracy of distinguishing between fatigue and alert states to enable reliable countermeasure devices. While physiological measurements like EEG are considered reliable for detecting fatigue, existing methods often rely on power spectral density (PSD) for feature extraction, which may not fully capture the complex temporal structures of brain signals. This paper proposes a novel pipeline combining Independent Component Analysis by Entropy Rate Bound Minimization (ICA-ERBM) for source separation, Autoregressive (AR) modeling for feature extraction, and a Bayesian neural network for classification. The experimental data comprised EEG recordings from 43 healthy participants aged 18 to 55, collected during a simulated driving task designed to induce fatigue. Participants performed a monotonous driving simulation until performance decrements or physiological signs of fatigue were observed. EEG signals were recorded using a 32-channel system and downsampled to 256 Hz. Data segments of 20 seconds were selected for both alert and fatigue states, validated through video monitoring, electrooculography (EOG), and subjective psychometric scales. The preprocessing pipeline utilized Second-Order Blind Identification (SOBI) and Canonical Correlation Analysis (CCA) to remove artifacts, followed by ICA-ERBM to separate independent neural sources. Features were extracted using AR modeling, which was compared against traditional PSD methods. The classification was performed using a Bayesian neural network, which optimizes network structure and prevents overfitting by incorporating hyper-parameters into the cost function. The results demonstrated that the proposed method significantly outperformed traditional approaches. The combination of ICA-ERBM, AR modeling, and the Bayesian neural network achieved a classification accuracy of 88.2%, with a sensitivity of 89.7% and a specificity of 86.8%. This configuration yielded an Area Under the Receiver Operating Characteristic curve (AUC-ROC) of 0.93, which was statistically superior (p < 0.05) to the AUC-ROC of 0.81 obtained when using PSD as the feature extractor without ICA-ERBM. The Bayesian framework allowed for the selection of an optimal network structure with six hidden nodes, providing robust generalization without requiring a separate validation set. The significance of this work lies in the validation of ICA-ERBM as an effective source separation technique for EEG-based fatigue detection, particularly when paired with AR modeling. The study confirms that exploiting both non-Gaussian properties and sample correlation in EEG signals enhances classification performance compared to standard frequency-domain methods. These findings suggest that the proposed system is a viable candidate for real-time driver fatigue monitoring systems, offering a robust and reliable solution to mitigate fatigue-related risks on the road.
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 | unpaywall | — | — | 2 | 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 | — | — | 16 | 2026-08-11 |
| verify | partial | — | — | — | 1 | 2026-08-10 |
Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified_with_issues.
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