Driver Distraction Detection Using Semi-Supervised Machine Learning

• Published in: IEEE transactions on intelligent transportation systems Vol. 17; no. 4; pp. 1108 - 1120
• Main Authors: Liu, Tianchi, Yang, Yan, Huang, Guang-Bin, Yeo, Yong Kiang, Lin, Zhiping
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.04.2016; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Advanced driver assistance system; Classification; Cost engineering; driver distraction; Driving conditions; eye movement; Eyes; Feature extraction; Intelligent transportation systems; Labeling; Learning; machine learning; Neural networks; Position measurement; Real-time systems; semi-supervised learning; Semisupervised learning; Support vector machines; Telematics; Training; Training data; Vehicles; Advanced driver assistance system; semi-supervised learning; machine learning; driver distraction; eye movement
• ISSN: 1524-9050; 1558-0016
• DOI: 10.1109/TITS.2015.2496157

Real-time driver distraction detection is the core to many distraction countermeasures and fundamental for constructing a driver-centered driver assistance system. While data-driven methods demonstrate promising detection performance, a particular challenge is how to reduce the considerable cost for collecting labeled data. This paper explored semi-supervised methods for driver distraction detection in real driving conditions to alleviate the cost of labeling training data. Laplacian support vector machine and semi-supervised extreme learning machine were evaluated using eye and head movements to classify two driver states: attentive and cognitively distracted. With the additional unlabeled data, the semi-supervised learning methods improved the detection performance (G-mean) by 0.0245, on average, over all subjects, as compared with the traditional supervised methods. As unlabeled training data can be collected from drivers' naturalistic driving records with little extra resource, semi-supervised methods, which utilize both labeled and unlabeled data, can enhance the efficiency of model development in terms of time and cost.