An Approach to Automatic Driver Sleep Detection

Abstract

This paper presents an approach to automatic sleep detection using electroencephalogram signals as a differentiating basis. The multi-layer perceptron and the support vector machine models are developed and analyzed with training and testing datasets. The results are evaluated using a cross-validation technique and compared with manual classification hypnogram tables. The models are very successful with sleep stage detection reaching up to 94%, and Cohen's index reaching up to 0.68, showing considerable promise for deployment and future studies.