Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM65670
Volume:
10 |
6 |
Jun 2026
Pages:
1-9
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.
Artificial Intelligence and Data Science
Zeal College of Engineering and Research
Driver fatigue, distraction, and abnormal driving behaviour remain leading contributors to road accidents worldwide, underlining the urgent need for intelligent monitoring systems that assess driver state from multiple complementary signals [1], [2]. This paper presents SafeRider, an AI-driven multimodal driver behaviour monitoring system that fuses a transfer-learned Convolutional Neural Network (CNN) image classifier with an ExtraTrees-based sensor classifier to predict the driver's state in real time. The visual stream is processed by a MobileNetV2 backbone with a custom multilabel head that independently predicts eye state (open/closed) and mouth state (yawning/neutral). In parallel, a 7-feature numerical payload — vehicle telemetry (speed, RPM, throttle, brake, steering angle) and physiological signals (heart rate, stress level) — is enriched with six engineered features and classified into five behavioural categories (attentive, distracted, drowsy, aggressive, eco-driving) using an ExtraTrees classifier trained on a SMOTE-balanced dataset. A rule-based fusion layer combines the two streams into a graded verdict (Driver Alert, Warning, Critical). Experimental results on a curated dataset of 301 facial images and 1,200 sensor rows demonstrate a multilabel CNN combined-accuracy of 96.05 % and a sensor classifier hold-out accuracy of 92.92 % (5-fold CV 94.48 %), validating that the proposed multimodal fusion approach improves reliability and reduces false predictions compared to single-model systems [8], [17], [19]. The framework is deployed as a production-style Flask service consumed by a React dashboard with text-to-speech alerts, providing an effective foundation for intelligent driver monitoring and advanced driver assistance systems.
KEYWORDS
This paper presents SafeRider, an AI-based hybrid system for real-time driver behaviour monitoring that fuses transfer-learned visual analysis with engineered-feature sensor analysis. The combination of a MobileNetV2 CNN with a multilabel head for facial cues and an SMOTE-balanced ExtraTrees classifier for vehicle and physiological data enables accurate detection of driver states such as attentive, drowsy, distracted, aggressive, and eco-driving. The deterministic fusion ladder improves accuracy and reduces false detections compared to single-modality baselines [17], [19], while remaining fully interpretable to safety reviewers. The Flask backend efficiently manages predictions with production-style hardening, and the React frontend provides a clear, accessible, and audibly alerting user interface. Experimental results achieve a CNN combined accuracy of 96.05 %, a sensor classifier hold-out accuracy of 92.92 %, and an end-to-end fusion accuracy of approximately 96 %, demonstrating the system's reliability and effectiveness for driver-safety applications.
[1] A. G. Wheaton, "Drowsy driving and risk behaviors—10 states and Puerto Rico 2011–2012," MMWR Morb Mortal Wkly Rep, vol. 63, no. 26, pp. 557–562, 2014. [2] NHTSA. Accessed: Feb. 27, 2024. [Online]. Available: https://www.nhtsa.gov/risky-driving/drowsy-driving [3] Canalys. Accessed: Feb. 20, 2024. [Online]. Available: https://www.canalys.com/newsroom/canalys-autonomous-driving-starts-to-hit-mainstream-as-35-million-new-cars-had-level-2-features-in-q4-2020 [4] O. Carsten, F. C. H. Lai, Y. Barnard, A. H. Jamson, and N. Merat, "Control task substitution in semiautomated driving: Does it matter what aspects are automated?" Hum. Factors, vol. 54, no. 5, pp. 747–761, Oct. 2012, doi: 10.1177/0018720812460246. [5] J. W. Jenness, N. D. Lerner, S. Mazor, J. Osberg, and S. Tefft, "Use of advanced in-vehicle technology by young and older early adopters. Survey results on adaptive cruise control systems," NHTSA, U.S. Dept. Transp., Washington, DC, USA, Tech. Rep. DOT HS 810 917, 2008. [6] IDTechEx. Accessed: Feb. 15, 2024. [Online]. Available: https://www.idtechex.com/ja/research-article/regulations-drivers-for-mandating-driver-monitoring-systems/30322 [7] U.S. Congress, S. 1406. Accessed: Feb. 15, 2024. [Online]. Available: https://www.congress.gov/bill/117th-congress/senate-bill/1406/text [8] M. Q. Khan and S. Lee, "A comprehensive survey of driving monitoring and assistance systems," Sensors, vol. 19, no. 11, p. 2574, Jun. 2019, doi: 10.3390/s19112574. [9] S. Titare, S. Chinchghare, and K. N. Hande, "Driver drowsiness detection and alert system," Int. J. Sci. Res. Comput. Sci., Eng. Inf. Technol., vol. 7, pp. 583–588, Jun. 2021, doi: 10.32628/cseit2173171. [10] C. Solomon and Z. Wang, "Driver attention and behavior detection with Kinect," J. Image Graph., vol. 3, no. 2, pp. 1–6, 2015.
Publication Date
6/29/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
0
Total views since publication
PDF Downloads
7
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...