Comparison of Lightweight Methods for Vehicle Dynamics-Based Driver Drowsiness Detection

2025-06-08

AI Summary

Driver drowsiness detection (DDD) methods driven by vehicle dynamics suffer from unreliable performance metrics, poor experimental reproducibility, and non-public datasets. Method: Building upon Aygun et al.’s publicly available dataset, this work establishes a transparent, fair, and data-leakage-proof standardized evaluation framework. It introduces a configurable, lightweight pipeline for feature extraction and modeling, enabling systematic comparison of classifiers—including random forest (RF), support vector machine (SVM), and k-nearest neighbors (k-NN). Contribution/Results: We identify significant reliability flaws in existing non-standardized approaches and empirically demonstrate RF’s superiority under resource constraints (88% accuracy). All code, configurations, and evaluation protocols are fully open-sourced to ensure complete reproducibility. Our analysis exposes two pervasive bottlenecks in the field: metric distortion and dataset opacity. This work proposes a new paradigm for trustworthy DDD evaluation, advancing methodological rigor and transparency in vehicular driver-state monitoring.

Technology Category

• Machine Learning: Evaluation and Analysis
• Computer Vision: Adversarial Attacks & Robustness
• Intelligent Robots: State Estimation

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Security and Privacy: Large-scale security measurements
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Driver drowsiness detection (DDD) prevents road accidents caused by driver fatigue. Vehicle dynamics-based DDD has been proposed as a method that is both economical and high performance. However, there are concerns about the reliability of performance metrics and the reproducibility of many of the existing methods. For instance, some previous studies seem to have a data leakage issue among training and test datasets, and many do not openly provide the datasets they used. To this end, this paper aims to compare the performance of representative vehicle dynamics-based DDD methods under a transparent and fair framework that uses a public dataset. We first develop a framework for extracting features from an open dataset by Aygun et al. and performing DDD with lightweight ML models; the framework is carefully designed to support a variety of onfigurations. Second, we implement three existing representative methods and a concise random forest (RF)-based method in the framework. Finally, we report the results of experiments to verify the reproducibility and clarify the performance of DDD based on common metrics. Among the evaluated methods, the RF-based method achieved the highest accuracy of 88 %. Our findings imply the issues inherent in DDD methods developed in a non-standard manner, and demonstrate a high performance method implemented appropriately.

Problem

• Compares vehicle dynamics-based drowsiness detection methods
• Addresses reliability and reproducibility issues in existing methods
• Evaluates performance using a transparent framework and public dataset

Innovation

• Uses public dataset for transparent comparison
• Implements lightweight ML models framework
• Random forest method achieves 88% accuracy