Creating Simplified Version of Lip Database based on Front View of Face

Abstract

Recently lot of work has been done on audio visual speech recognition but less work has been done on visual speech and speaker recognition. This research belongs to human computer interaction (HCI) domain. HCI makes human computer interaction simple. This paper represents the creating of database of visual speech and speaker in English language and preprocessing of it to improve recognition accuracy. We have studied Tulipse1 database, AV Database and CUAVE Database on the basis of these different databases we have created our own database. This is useful for all researchers those are working HCI domain particularly Visual Speech and Speaker Recognition.