Development of a humanoid robot mouth with text-to-speech ability

Abstract

This is a study of lip movements when speaking the English language, developed humanoid robot face with moving lips and text-to-speech ability. A humanoid robot face is a robotic face capable of mimicking an actual human being during speech. Humanoid robots are fnding their way into human activities rapidly with the advancement of technology. Humanoid robots will become very close to humans since they can produce human-like speech by synchronized realistic lip movements while providing sound. However, most humanoid robots do not have very much correct lip movements at the moment. A robotic face with more realistic lip movements will apply in several situations such as news reading, teaching, storytelling, interactive robotic machines, and live advertisements. During the development, a robotic model was built that could simulate jaw movements. Additional lip features were added with driver mechanisms to move the lips in synchronization with the voice produced; audio and text analyzing technology was utilized to improve the accuracy based on phonetic symbols of the English language. Movements of the tongue are essential when reads phonetics in the English language. In this development, physical equipment does not generate vocals; therefore, we neglected tongue movements and facial expressions. We only focus on accurate synchronized lip movements using three control points. This analysis can be used to simulate lip movements in the model of a humanoid robotic face. The performance was tested, and the closeness of the robot-generated lip movements to actual human lip movement was improved by using many voice samples and digitized lip coordinates with the help of image processing techniques.