· Contributors · Organizations ·
Affect-enhancing speech features for robotic communication
Human AI Robot Teaming (HART)
DescriptionAccording to the theory of mind, the perception of experience and agency determines how robots are treated - with higher experience leading to more sympathy and forgiveness with robots. One strategy to increase the experience ratings of robots while not increasing agency is the application of affective communication. The two components of speech that have been considered for the robotic communication in this study are the prosodic and the verbal component. In our online study, 30 participants listened to audios in which robots introduced themselves. The synthesized speech was created by text-to-speech software once with and once without prosody. The verbal content differed in regard to emotional words and subjective narration. Results showed that the use of emotional words increased the experience ratings of robots whereas agency was not affected. Robots speaking with prosody showed a tendency to enhance the attribution of feelings. Only speaking in a more subjective way had no impact on the experience dimension. Results can be used to improve the emotional impact of robotic speech and thus increase the acceptance of robots. However, further studies should control for different robot types as well as contextual influences and replicate findings in laboratory settings.