The development of a music reading and singing two‐wheeled robot

Abstract The objective of this research was to construct a two‐wheeled robot that can autonomously read music and sing songs with a synthesized voice. The music for the robot was created by using a musical notation editor developed in this study, which could be read by the robot's vision system. A musical notation recognition program converted the image obtained from the vision system to a text file, and then a voice synthesis program processed the text file to generate the synthesized music autonomously. We tested four Mandarin songs and the accuracy of recognition was over 98%. It demonstrated that a robot with integrated functions should be promising in entertainment applications in the future.