Drowsiness Detection using EEG signals and Machine Learning Algorithms
DOI: 10.1051/itmconf/20224403030
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This paper addresses the critical safety issue of driver drowsiness, a leading cause of vehicular accidents characterized by reduced vigilance and delayed reaction times. While existing detection systems rely on video-based monitoring of physical features like eye closure or steering wheel movement, these methods are often limited by environmental factors such as lighting conditions and driver distraction. To overcome these limitations, the authors propose an electroencephalogram (EEG)-based detection system, leveraging EEG’s high temporal resolution and non-invasive nature to directly monitor physiological states of alertness. The study specifically investigates the efficacy of deep learning architectures compared to traditional machine learning algorithms for classifying EEG signals into drowsy and awake states. The experimental design utilizes the "EEG Eye State Dataset" from the UCI Machine Learning Repository, which contains 14,304 samples recorded over 117 seconds using an Emotiv EEG Neuroheadset with 14 electrodes. The dataset is randomly split into 80% for training and 20% for testing. The authors compare two distinct methodological approaches. The first involves traditional machine learning classifiers—Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Logistic Regression (LR), and Naïve Bayes—which require manual feature extraction. For these models, statistical features including mean, standard deviation, kurtosis, skewness, and Discrete Cosine Transform (DCT) coefficients were extracted from the raw EEG signals. The second approach employs a Convolutional Neural Network (CNN), a deep learning model that performs automatic feature extraction. The CNN architecture consists of two convolutional layers followed by three fully connected dense layers, utilizing ReLU activation for hidden layers and sigmoid for the output layer, with dropout regularization applied to prevent overfitting. The results demonstrate a significant performance disparity between the two approaches. The CNN model achieved the highest accuracy of 94.75%, with a precision of 93.6%, recall of 95.7%, and an F1-score of 94.6%. In contrast, the traditional machine learning models yielded lower performance metrics: SVM achieved 83.7% accuracy, KNN reached 78.6%, Logistic Regression scored 77.3%, and Naïve Bayes performed the poorest at 59.4%. The superior performance of the CNN is attributed to its ability to automatically generate invariant features through convolution and pooling operations, eliminating the need for manual feature engineering and reducing the risk of human error in feature selection. The study concludes that deep learning, specifically CNNs, offers a more robust and accurate solution for EEG-based drowsiness detection compared to conventional machine learning techniques. By delegating feature extraction to the neural network, the proposed method achieves higher detection capability, suggesting its viability for developing reliable, real-time driver monitoring systems. The findings highlight the potential of integrating deep learning with physiological signal processing to enhance road safety by providing more accurate and less intrusive drowsiness 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 | unpaywall | — | — | 2 | 2026-08-09 |
| extract | success | cached | — | — | 124 | 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 |
| promote | success | — | — | — | 1 | 2026-08-09 |
| summarize | success | llm | qwen3.6-27b-nvidia | summ-v5 | 123 | 2026-08-10 |
| tag | success | vector_similarity | — | — | 11 | 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