Prosodic alignment in human-computer interaction

Androids replicating humans in forms need also to achieve replication in behaviors to realize high level of believability. It has been observed that people exhibit strong tendency to adjust to each other a number of speech and language features in human-human conversational interactions to obtain communication efficiency and emotional engagement. We investigate in this paper the phenomena of mutual alignment of speech characteristics in human-computer interactions, with particular focus on human to computer alignment of speech prosody features. We found that people exhibit one directional spontaneous short-term alignment of loudness and switching pause durations of their speech to computer produced speech, even without the visual presence of personified characters. We believe this phenomena of prosodic adaptation provides one of the key components for building empathy between humans and androids.