Javascript must be enabled for the correct page display

Multimodal affective human-machine interaction

Verhoef, T. (2008) Multimodal affective human-machine interaction. Master's Thesis / Essay, Artificial Intelligence.

[img]
Preview
Text
AI-MAI-2008-T.Verhoef.pdf - Published Version

Download (16MB) | Preview

Abstract

Present research in Artificial Intelligence brings us ever closer to a reality in which machines and robots assist humans in their everyday lives. These machines will be tutoring us, be concerned with our health, taking care of our elderly and play the part of a companion in our homes. This development increases the need for good human-machine interaction. The field of Affective Computing addresses the importance of emotions in human-machine interaction. This thesis aims to provide an overview of this field and to provide possible solutions for the design and implementation of an affect-sensitive empathetic agent that can sense and interpret the affective state of a user (affect recognition) and reacts appropriately by showing an empathetic response (affect generation). A data set was created by exposing participants to emotion eliciting stimuli while their physiological signals and facial expressions were recorded. This data was used to train and test several emotion recognizers, using both discrete and continuous emotion representations. Continuous classification appeared to be more successful than the use of discrete categories. In addition, an anthropomorphic avatar was implemented to express twelve different psychologically grounded facial expressions. These expressions were tested for recognizability and believability in a small user study. The success rate differed for each expression, but all twelve were recognized with a percentage above chance level.

Item Type: Thesis (Master's Thesis / Essay)
Degree programme: Artificial Intelligence
Thesis type: Master's Thesis / Essay
Language: English
Date Deposited: 15 Feb 2018 07:28
Last Modified: 15 Feb 2018 07:28
URI: http://fse.studenttheses.ub.rug.nl/id/eprint/8473

Actions (login required)

View Item View Item