Key points are not available for this paper at this time.
As social robots become more integrated into daily life, understanding human perceptions of them is essential for effective design. This mixed factorial study examined how four emotions (happiness, sadness, anger, and relief) and a humanoid robot’s semantic-free utterances (SFUs) with male or female voices influence user perception. Forty-nine young participants listened to a fairy tale while the humanoid robot Pepper responded with gestures and emotional non-speech sounds. Emotion recognition accuracy and social perception were significantly higher in the Voice + Gesture condition than in the Gesture-Only condition. Happiness and sadness were most accurately recognized. Male voices elicited greater discomfort and perceptions of anthropomorphism, while male participants experienced more discomfort overall and female participants felt warmer toward Pepper. However, voice type and participant gender did not affect emotion recognition accuracy. The results provide insights for developing more effective and emotionally expressive social robots.
Liu et al. (Thu,) studied this question.