Effect of facial colors on humanoids in emotion recognition using speech

摘要

The purpose of This work is to measure the effect of facial colors on a humanoid robot in emotion recognition. Several people were asked to evaluate the effect on emotion of the robot by ignoring or including facial colors together with natural speech or machine generated speech. A multimodal expression system used for the evaluation consisted of two subsystems-a facial expression subsystem having the ability of expressing three facial colors based on color psychology, that is, yellow (expressing joy), red (anger) and blue (sadness), and a speech mimicking subsystem that resynthesizes emotional speech made by a human based on prosody extraction of pitch, loudness and tempo. Our experiments show that the emotion recognition rate increases by 25% and the certainty of recognition improves when using these facial colors.