Davide Conflitti, Riccardo Santoboni, Carlo Giovannella, Andrea Paoloni
pp. 107-110 – download
(https://doi.org/10.55612/s-5002-005_6-017)
Abstract
We present a new tool and test procedure – based on the use of the finite state emotional model described by the Plutchick ‘flower’ – designed to measure the ability of the human to recognize the emotion carried by the voice when the italian language is used. Using the no-sense sentences contained in the corpus EMOVO we have been able to that in natural conditions the human ability to recognize the emotional states using only the vocal channel are quite poor (30% in average). The recognition rate depends on many factors, among them the sex and the peculiarities of the speaker, the specific emotion, the specific contest.
keywords: Emotivity, affective computing, emotional colored voice corpora, multimodal interaction, human factors evaluation