Large Sleepy Reading Corpus (LSRC): Applying Read Speech for Detecting Sleepiness

摘要

This paper describes a Large Sleepy Reading Corpus (LSRC) based on sleep deprivation data (N=402; total duration 22 h). During the sleep deprivation, a standardized self-report scale was used just before the recordings to determine the sleepiness state. The speech material consisted of different reading passages. In order to investigate sleepiness induced speech changes, a standard set of spectral and prosodic features was extracted from recordings. After applying a standard openSMILE feature set, and a SVM regression we achieved correlation coefficients of .44 for male and .53 for female speaker.