Perception of visual scene and intonation patterns of robot utterances

Abstract

Assigning intonation to dialogue system output in a way that reflects relationships between entities in the discourse context can enhance the acceptability of system utterances. Previous research concentrated on the role of linguistic context; dialogue situatedness and the role of visual context in determining accent placement has not been studied. We present an experimental study on the influence of visual context on the perception of nuclear accent placement in synthesized clarification requests. We found that utterances are perceived as appropriate more often when the visual scene licenses the nuclear accent placement than when it does not.