Contrastive accents : how to get them and what do they look like

Abstract

Automatic dialog systems tested with naive users are often confronted with special speaking styles, as e.g. words produced with emphatic or contrastive accent. Such utterances usually cause problems for word recognizers, because they were not included in the training data. It is thus important for the improvement of future systems to be able to collect utterances containing contrastive accents produced as natural as possible. We describe in this paper an automatic simulation system for provoking and collecting contrastive accents. With this system, 15 recording sessions were conducted, in total 205 word tokens produced either with default or with contrastive accent were collected. We discuss the results of an automatic classification as well as the relevance of extracted prosodic features for the marking of contrastive accents.