Abstract
Drowsiness is a major cause of fatal accidents, in particular in transportation. It is therefore crucial to develop automatic, real-time drowsiness characterization systems designed to issue accurate and timely warnings of drowsiness to the driver. In practice, the least intrusive, physiology-based approach is to remotely monitor, via cameras, facial expressions indicative of drowsiness such as slow and long eye closures. Since the system’s decisions are based upon facial expressions in a given time window, there exists a trade-off between accuracy (best achieved with long windows, i.e., at long timescales) and responsiveness (best achieved with short windows, i.e., at short timescales). To deal with this trade-off, we develop a multi-timescale drowsiness characterization system composed of four binary drowsiness classifiers operating at four distinct timescales (5 s, 15 s, 30 s, and 60 s) and trained jointly. We introduce a multi-timescale ground truth of drowsiness, based on the reaction times (RTs) performed during standard Psychomotor Vigilance Tasks (PVTs), that strategically enables our system to characterize drowsiness with diverse trade-offs between accuracy and responsiveness. We evaluated our system on 29 subjects via leave-one-subject-out cross-validation and obtained strong results, i.e., global accuracies of 70%, 85%, 89%, and 94% for the four classifiers operating at increasing timescales, respectively.
Highlights
Drowsiness is defined as the intermediate, physiological state between wakefulness and sleep.It is associated with a difficulty to stay awake, a strong desire to fall asleep, and is characterized by impairments of performance, both cognitive [1,2] and motor [3,4]
We present a multi-timescale system to deal with the trade-off between accuracy and responsiveness; we introduce an appropriate multi-timescale ground truth to train such a multi-timescale system, which is based on objective, performance-based indicators, i.e., the reaction times (RTs) performed during Psychomotor Vigilance Tasks (PVTs); we use the sequence of raw eyelids distances as the intermediate representation, which we show to lead to strong results when processed by a multi-timescale temporal convolutional neural networks (CNNs); we adopt a strict, rigorous evaluation scheme, and compare, by proxy, the performance of our system with the performances of systems of other studies; we make our drowsiness dataset, code, and trained models available
We have presented a new multi-timescale drowsiness characterization system that is novel, data-driven, automatic, real-time, and generic
Summary
Drowsiness is defined as the intermediate, physiological state between wakefulness and sleep. We present a multi-timescale system to deal with the trade-off between accuracy and responsiveness; we introduce an appropriate multi-timescale ground truth to train such a multi-timescale system, which is based on objective, performance-based indicators, i.e., the RTs performed during PVTs; we use the sequence of raw eyelids distances (produced by a CNN, trained from scratch) as the intermediate representation, which we show to lead to strong results when processed by a multi-timescale temporal CNN; we adopt a strict, rigorous evaluation scheme (i.e., leave-one-subject-out cross-validation), and compare, by proxy, the performance of our system with the performances of systems of other studies; we make our drowsiness dataset, code, and trained models available (see details in Appendix)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
