@article{ref1,
title="Analysis of drivers' behavior before and after crashes in simulated expressway driving to predict sleepiness levels for doze alarm activation",
journal="JSAE review",
year="1998",
author="Numata, N. and Kitajima, H and Goi, Y and Yamamoto, Koichi",
volume="19",
number="3",
pages="269-275",
abstract="Sleepiness related behavior was analyzed in a simulated expressway. Long blinking and decrease of eye movement were observed in 9 out of 10 subjects. Prediction of sleepiness change by these characteristics was examined. Rate of long blinking accounted generally for changes of drivers' sleepiness (r=0.71). Furthermore, mean correlation coefficient increased by using standard deviation of horizontal eye movement in combination with rate of long blinking (r=0.75). Actually, a passenger can evaluate driver's sleepiness based on these behavioral characteristics synthetically. In the same way, the accuracy of sleepiness prediction was increased by using several indices.<p />",
language="",
issn="0389-4304",
doi="10.1016/S0389-4304(98)00011-3",
url="http://dx.doi.org/10.1016/S0389-4304(98)00011-3"
}