Lip Reading Method를 이용한 자동차내 음성인식시스템

Abstract

In this paper, we use image and speech integrated recognition system for the convenience and safety of drivers in order to control equipments of inside of car. We make DB, constructed on AVI files, for the image and speech recognition. We get features form the movement of lip using Lip Reading Method in image recognition and one from the MFCC(Mel Frequency Cepstral Coefficients) in the part of speech recognition. We integrate using these features which will be applied to HTK recognition engine we construct the image and speech recognition system. We show that wrong recognition results, induced by noise environment in speech recognition system can be corrected by image recognition system.