The influence of vocal effort on human speaker identification

Abstract

Although many of the acoustic cues used for speaker identification change systematically with the voice level of the talker, little is known about the influence that vocal effort has on the identification of individual talkers by human listeners. In this experiment, listeners were trained to identify four different same-sex talkers speaking at one of three different levels of vocal effort (whispered, conversational, or shouted). They were then tested on their ability to identify the same four talkers speaking at the other two levels of vocal effort. The results show that the whispering talkers were substantially harder to identify than the conversational talkers, and that the conversational talkers were substantially harder to identify than the shouting talkers. The results also show that listeners who were trained to identify individual talkers speaking at one level of vocal effort had difficulty identifying the same talkers when they were speaking at a different level of vocal effort. These results confirm that changes in vocal effort have a dramatic impact on human speaker identification, and suggest that the effects of vocal effort should be considered in the design of automatic speaker identification systems.