Expressive humanoid face: A preliminary validation study

Non-verbal signals expressed through body language play a crucial role in our daily communications. Facial expressions, in particular, are the most universal signs to express innate emotional cues. Human faces convey important information in social interactions, which help us to better understand our interlocutor. Nowadays, humanoids and social robots are becoming increasingly similar to humans both aesthetically and expressively. However, their visual expressiveness is a crucial issue in making these robots more realistic and intuitively perceived as human-like. This paper presents a preliminary study aimed at evaluating the capability of a humanoid to perform facial expressions in terms of recognition rate and response time in comparison with humans' ability. Results showed that the recognition rate of human and robot expressions did not reveal differences while the physical robot can convey expressions better than its 2D photos and its 3D models. Moreover, the results showed that both human and robot positive expressions were better recognized than the negative ones.