Augmenting the audio-based expression modality of a non-affective robot | IEEE Conference Publication | IEEE Xplore