Virtual conversation with a real talking head

Abstract

A talking head is system performing an animated face model synchronized with a speech synthesis module. It is used as a presentation layer of a conversational agent which provide an answer. It provides an answer when a query is written as an input by the user. The textual answer is converted into facial movements of a 3D face model whose lips and tongue movements are synchronized with the sound of the synthetic voice. The client-server paradigm has been used for the WEB infrastructure delegating the animation and synchronization to the client, so that the server can satisfy multiple requests from clients; while the Chatbot, the digital signal processing and the natural language processing are provided by the server.