A novel drowsiness detection scheme based on speech analysis with validation using simultaneous EEG recordings

This paper uses voice response analysis of human subjects for assessing their level of fatigue. The results are simultaneously validated through Electroencephalography (EEG) based measurements. We have designed a 36-hour long experiment where the subjects are asked to repeat a particular sentence at different stages. The response is analyzed for computing various parameters such as voiced duration, unvoiced duration, and the response time. We have used Mel-Frequency-Cepstral-Coefficients (MFCC) as the features for the silence, voiced and unvoiced parts of speech. We have segregated these parts using a Gaussian Mixture Model (GMM) classifier. The results have been validated with an EEG based parameter i.e. relative energy of α band which increases with fatigue. A correlation between Speech and EEG based measurements is observed at various stages of the experiment.