EEG-based Safety Driving Performance Estimation and Alertness Using Support Vector Machine
DOI: 10.14257/ijsia.2015.9.6.13
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 drowsiness, a leading cause of traffic accidents, by developing an EEG-based Safety Driving Performance Estimation and Alertness (SDPEA) system. The authors aim to create a robust, non-invasive method for monitoring driver cognitive states in real-time by analyzing brain activity changes during simulated driving. The motivation stems from the need for more accurate detection methods than traditional physiological measures like eye blinking or heart rate, leveraging the strong correlation between specific EEG frequency bands and fatigue levels. The experimental design involved nine healthy male subjects aged 21–30 performing a sustained-attention driving task in a simulator. To induce drowsiness, the 90-minute sessions were conducted in the early afternoon (13:00–14:00), a period statistically associated with sleepiness. Data acquisition utilized a 64-channel high-resolution EEG system sampled at 1000 Hz, with electrodes positioned according to the 10-20 international standard. Ground truth labels for wakefulness and sleepiness were established through manual annotation combining subject feedback and video recordings. The methodology employed a multi-stage processing pipeline: first, Independent Component Analysis (ICA) was applied to remove artifacts such as eye movements and muscle activity. Next, Fast Fourier Transform (FFT) was used to extract eight specific EEG-band power features: delta, theta, alpha, beta, and four derived ratios including (alpha+theta)/beta and theta/beta. Fisher score analysis was then utilized to select the most discriminative features, reducing redundancy. Finally, a Support Vector Machine (SVM) classifier with a radial basis function kernel was trained to distinguish between alertness levels. The results demonstrated that the SDPEA system could effectively estimate quantitative driving performance and alertness levels. Classification accuracy varied across subjects and features, with alpha, beta, (alpha+theta)/beta, and alpha/beta features yielding superior performance compared to others. For instance, subject S2 achieved 100% accuracy using alpha, beta, and alpha/beta features, while subject S4 showed lower accuracy (64.7%) with delta features. The system successfully categorized alertness into four states: wakefulness, two intermediate states, and sleepiness. Visual analysis of scalp topographies confirmed distinct spectral changes between wakefulness and sleepiness. However, the authors noted that band powers are susceptible to artifact contamination, emphasizing the necessity of rigorous preprocessing. The study concludes that EEG-based monitoring using ICA for artifact removal and SVM for classification is a viable approach for real-time drowsiness detection. The findings support the use of specific EEG band power ratios as robust indicators of cognitive state. The authors acknowledge limitations, noting that traditional machine learning methods may struggle with feature selection involving non-workload components and potential inaccuracies in predefined labels. They suggest that future work should focus on improved mathematical algorithms for workload quantification to enhance the reliability of alertness assessment systems.
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 | canonical_url | — | — | 1 | 2026-08-09 |
| extract | success | cached | — | — | 3 | 2026-08-10 |
| clean | success | clean | — | — | 1 | 2026-08-09 |
| chunk | success | chunk | — | — | 1 | 2026-08-09 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-08-09 |
| enrich | failed | — | — | — | 1 | 2026-08-09 |
| promote | success | — | — | — | 1 | 2026-08-09 |
| summarize | success | llm | qwen3.6-27b-nvidia | summ-v5 | 2 | 2026-08-10 |
| tag | success | vector_similarity | — | — | 10 | 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