Presented by Gordon McIntyre at the Eleventh Australasian
International Conference on Speech Science and Technology SST2006,
Auckland, New Zealand, 6-8 December 2006
Abstract
Many studies into affective computing cover similar ground or overlap,
pointing to much duplication of effort. Two problem areas exist in the
research. The first relates to the way in which emotions are defined
and described. The second is in the techniques used to elicit samples
of speech. Further, few studies are based on natural speech, ignoring
social, contextual, cultural and agent characteristics and few take
account of the affective state of the receiver. This paper presents a
survey of the problem areas and proposes 1) a comprehensive model of
affective communication; and 2) a set of ontologies of affective
computing.
AUTHOR = {G. McIntyre and R. Goecke},
TITLE = {{Researching Emotions in Speech}},
BOOKTITLE = {{Proceedings of the Eleventh Australasian International Conference on Speech Science and Technology SST2006}},
PUBLISHER = {ASSTA},
ADDRESS = {Auckland, New Zealand},
PAGES = {264--369},
MONTH = dec,
YEAR = 2006}