Citation

Ahmad, Khubab and Em, Poh Ping and Ab Aziz, Nor Azlina (2025) Enhancing Driver Drowsiness Detection Using OBD-II and MFCC Features: A Transformer Based Approach. In: 2025 Multimedia University Engineering Conference, MECON 2025, 21 July 2025 - 23 July 2025, Cyberjaya, Malaysia.

Text 72.pdf - Published Version Restricted to Repository staff only Download (1MB)

Abstract

Driver drowsiness is a critical factor contributing to road accidents, demanding the development of accurate and responsive detection systems for intelligent transportation. This paper proposes a novel deep learning framework, DDD-GCViT, which combines On-Board Diagnostic II (OBD-II) sensor data and Mel-Frequency Cepstral Coefficients with a transformer-based architecture, the Global Context Vision Transformer, to detect drowsy driving states. Key vehicle parameters, including speed, throttle position, and steering torque, are extracted from OBD-II logs and transformed into MFCC features to capture temporal and frequency characteristics of driving behaviour. Ground truth labels for driver alertness are generated through synchronized camera recordings and analysed using a pretrained vision model. These labels are aligned with the MFCC features to form structured image-like representations, serving as input to the proposed model. Unlike conventional convolutional neural networks, the transformer-based architecture leverages global attention mechanisms to capture long-range dependencies, resulting in improved feature representation and classification performance. Experimental results demonstrate that the proposed model achieves a classification accuracy of 97.90%, outperforming several baseline CNN-based models. Evaluation metrics such as precision (97.90%), recall (97.91%), and F1- score (97.90%) further confirm the robustness and consistency of the approach. The proposed framework exhibits strong potential for real-time deployment in intelligent driver assistance systems. Its high accuracy, scalability, and effective use of multimodal data make it a promising solution for mitigating drowsiness-related incidents and enhancing road safety.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	OBD-II sensor data, global context vision transformer
Subjects:	T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL1-484 Motor vehicles. Cycles
Divisions:	Faculty of Engineering and Technology (FET)
Depositing User:	Ms Rosnani Abd Wahab
Date Deposited:	18 Mar 2026 07:45
Last Modified:	18 Mar 2026 07:45
URII:	http://shdl.mmu.edu.my/id/eprint/15561

Downloads

Edit (login required)