Disfluencies in change detection in natural, vocoded and synthetic speech
Rasmus Dall, Mirjam Wester, Martin M. B. Corley · 2015
In this paper, we investigate the effect of filled pauses, a discourse marker and silent pauses in a change detection experiment in natural, vocoded and synthetic speech. In natural speech change detec-tion has been found to increase in the presence of filled pauses, we extend this work by replicating ear-lier findings and explore the effect of a discourse marker, like, and silent pauses. Furthermore we re-port how the use of "unnatural " speech, namely syn-thetic and vocoded, affects change detection rates. It was found that the filled pauses, the discourse marker and silent pauses all increase change de-tection rates in natural speech, however in neither synthetic nor vocoded speech did this effect ap-pear. Rather, change detection rates decreased in both types of "unnatural " speech compared to nat-ural speech. The natural results suggests that while each type of pause increase detection rates, the type of pause may have a further effect. The "unnatural" results suggest that it is not the full pipeline of syn-thetic speech that causes the degradation, but rather that something in the pre-processing, i.e. vocoding, of the speech database limits the resulting synthesis.