A preliminary study of an audio-visual speech coder: Using video parameters to reduce an LPC vocoder bit rate

Today there exists numerous speech coding techniques which allow to transmit and stock efficiently the acoustic signals. But speech is both auditory and visual : audio and video information are complementary and the lip shape of the speaker can help the listeners to better understand what is said. This paper aims to show the interest of video information in the speech coding domain in general and in terms of transmission rate in particular.