Oratory or the art of public speaking with eloquence has been cultivated since ancient times. However, the fear of speaking in public-a disproportionate reaction to the threatening situation of facing an audience-affects a very important part of the population. This work arises from the need to help alleviate this fear through a tool where to train the ability of public speaking. To this purpose, we built a virtual reality system that offers the speaker a safe environment to practice presentations. Since the audience is the only way to receive feedback when giving a speech, our system offer s a virtual audience that reacts and gives real-time feedback based on the emotions conveyed by three parameters: voice tone, speech content and speaker's gaze. In this paper, we detail the modelling of a behavioural-realistic audience just focusing on the speakers' voice tone: 1) by presenting an algorithm that controls the audience' reactions based on the emotions beamed by the speaker, and 2) by carrying out an experiment comparing the reactions generated by the agents with those of a real audience to the same speech, in order to refine the given algorithm. In this experiment, the audience subjects are asked to fill a questionnaire-level of engagement and perceived emotions-for a speech performed by professional actors representing different emotions. Afterwards, we compared the reactions of said audience with the ones generated by our algorithm, and used the results to improve it.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.