Emotional temperature

Automatic emotional state recognition from the speech signal represents a remarkable improvement in human-machine interfaces and it opens up a wide range of new applications. This turns out to be no trivial task due to the degree of difficulty inherent in the study of emotions. Traditional methods of emotional discrimination use prosodic and paralinguistic features, which are determined by a linguistic segmentation of the locution. This type of segmentation results in almost real scenarios impossible to estimate. In this paper a simple and effective method of automatic discrimination between positive and negative emotional intensity speech is presented. This work proposes a new strategy based on a few features set obtained from a temporal segmentation of the speech signal. This strategy is robust, offers low computational cost and improves the performance of a segmentation based on linguistic aspects.