A Decade of Progress in Drowsiness Detection Using Facial Features
A Comprehensive Bibliometric Analysis and Thematic Mapping
Keywords:
Advanced driver assistance systems, bibliometric analysis, computer vision, convolutional neural networks, deep learning, drowsiness detection, driver monitoring, eye tracking, facial features, Internet of Things
Abstract
This study presents a comprehensive bibliometric analysis of research trends in drowsiness detection using facial features, examining publications from 2014 to 2024. Employing a PRISMA-guided methodology, we extracted and analyzed 347 documents from the Scopus database, revealing significant patterns in research productivity, citation impact, geographical distribution, institutional contributions, and thematic evolution. Our analysis identifies an annual growth rate of 18.59% in publication volume, with a notable surge after 2019, coinciding with the widespread adoption of deep learning approaches. Geographically, Asian countries dominate research output, with India (182 publications) and China (100 publications) leading contributions, while China garnered the highest citation impact (1370 citations). Through sophisticated co-occurrence network analysis, we identified four distinct research clusters: (1) Physiological Insights via Neural Networks, (2) Computer Vision-Based Drowsiness Detection, (3) Multi-Modal Fatigue Detection for Accident Prevention, and (4) Deep Learning for Biomedical Signal Analysis. Temporal analysis of keyword evolution reveals a shift from traditional machine learning approaches toward deep neural networks, Internet of Things integration, and real-time monitoring systems. Our thematic mapping further categorizes research into basic themes (CNNs, eye aspect ratio), motor themes (driver fatigue detection, cloud computing), niche themes (3D head pose estimation, behavioral measurement), and emerging/declining themes (pupil detection, blink detection systems). Systematic analysis of deployment metrics reveals critical gaps: only 3.3% of top-cited papers report frames per second (FPS), and none report latency or time-to-alarm (TTA), despite frequent claims of real-time capability. Analysis of multimodal approaches shows 74% of studies focus exclusively on facial features, while 26% incorporate physiological or vehicular signals. This comprehensive bibliometric landscape illuminates the field's evolution, identifies research gaps particularly regarding deployment readiness, ethnicity-specific considerations and low-light environments, and provides a strategic roadmap for future research directions in drowsiness detection systems.
Published
2026-02-05
How to Cite
Subowo, M. H., Andono, P. N., Shidik, G. F., & Santoso, H. A. (2026). A Decade of Progress in Drowsiness Detection Using Facial Features. Statistics, Optimization & Information Computing. https://doi.org/10.19139/soic-2310-5070-3192
Issue
Section
Research Articles
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
