A Review on Deep Learning Techniques for EEG-Based Driver Drowsiness detection systems
archive: archived pipeline: cataloged
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This review paper addresses the critical safety issue of driver drowsiness, a leading cause of road accidents and fatalities, by evaluating the efficacy of deep learning (DL) techniques for detecting drowsiness using electroencephalogram (EEG) signals. The authors note that while traditional machine learning methods have been widely used, they suffer from limitations such as the need for massive datasets and hand-crafted feature extraction. Consequently, the research focuses on DL approaches, which offer superior performance in physiological signal classification by automatically learning features. The study aims to synthesize recent literature (primarily from the last three years), outline the standard pipeline for EEG-based detection systems, and identify current limitations and future directions. The methodology involves a systematic review of recent journal and conference articles selected via specific keywords related to "deep learning," "EEG," and "driver drowsiness." The paper details the standard architecture of these systems, which includes data acquisition via scalp electrodes (often using the international 10-20 system), preprocessing steps (artifact removal, normalization, and feature extraction via Fourier or Wavelet transforms), and classification using DL models. The review categorizes DL techniques into Convolutional Neural Networks (CNNs), including 1D and 2D variants, and Recurrent Neural Networks (RNNs), specifically Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU). It also catalogs publicly available datasets such as Sleep-EDF, SEED-VIG, and MIT/BIH Polysomnographic EEG, noting their subject counts, electrode configurations, and sampling rates. Key findings indicate that CNNs are the most prevalent and effective technique for this task, outperforming traditional ML methods. Specific studies reviewed report high accuracy rates; for instance, a Cascaded CNN with an attention mechanism achieved 97.26% accuracy, while a spatial-temporal CNN (ESTCNN) reached 97.37% accuracy. Other models, such as modified AlexNet and Inception-based architectures, reported accuracies ranging from 94% to 95.7%. The review highlights that 1D-CNNs offer lower computational complexity than 2D-CNNs, making them suitable for real-time applications, while LSTM and GRU models effectively handle the sequential nature of EEG time-series data. A significant trend observed is the shift toward using fewer electrodes (single or few-channel EEG) to improve usability and reduce cost, with studies showing that specific channels like FP1, T3, and P4 are highly informative for drowsiness detection. The significance of this work lies in its comprehensive mapping of the state-of-the-art in EEG-based drowsiness detection. It concludes that while current systems achieve high precision, a major challenge remains the development of models that are both accurate and resource-efficient, capable of running on minimal spatial memory and processing power for real-time deployment. The authors propose that future research should focus on generalizing results across different subjects and environments, optimizing model size for mobile integration, and addressing the variability in EEG signal quality to create robust, early-warning systems that can effectively prevent drowsy-driving accidents.
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 | — | — | 5 | 2026-08-23 |
| 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 | qwen3.8-27b-gittensor | summ-v5 | 3 | 2026-08-23 |
| tag | success | vector_similarity | — | — | 17 | 2026-08-11 |
| verify | success | — | — | — | 1 | 2026-08-09 |
Summary generated by qwen3.8-27b-gittensor on 2026-08-23; verification: pending re-verification.
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