Animating TTS Messages in Android using Open-Source Tools

摘要

We describe in this paper how to use open-source resources to design and implement an Android application that renders a three-dimensional model of a human head to animate the lip movements of human speech from input text. The application utilizes the Android Text-To-Speech (TTS) engine[1] to convert any input text, which can be entered by the user in a text box or chosen from a menu of predefined messages, to human speech in English. Animation of the speech is carried out by a 3D graphics model of a human head composed of polygon meshes[20]. Blender[7, 30], a popular open-source graphics suite, is employed to create the 3D model and save its mesh data in the COLLAborative Design Activity (COLLADA) format[22], which is also an open graphics format. We use Java language to develop a parser[25, 42] to extract coordinates of polygons from a COLLADA file and organize the data into a format that can be rendered effectively by OpenGL ES, the graphics rendering library used by Android. The producer-consumer paradigm is employed to synchronize the animated lip movements and the speech generated by TTS. When the application is lying idle, it moves the head randomly to simulate other facial expressions such as blinking the eyes and yawning.