Lip reading from scale-space measurements
نویسندگان
چکیده
Systems that attempt to recover the spoken word from image sequences usually require complicated models of the mouth and its motions. Here we describe a new approach based on a fast mathematical morphology transform called the sieve. We form statistics of scale measurements in one and two dimensions and these are used as a feature vector for standard Hidden Markov Models (HMMs).
منابع مشابه
لبخوانی و ادراک گفتار دانشآموزان کمشنوای مدارس ویژۀ کمشنوایان در شهر تهران
Objective: The goal of this study was to evaluate the lip reading ability and Speech perception of hearing impaired students of special schools for the hearing impaired in different speech levels. Materials & Methods: In this cross- sectional study, 44 deaf students (9-12 years old) were selected with multi-stage cluster sampling method, from two special schools for the deaf in Tehran. Tools...
متن کاملImproving Lip-reading with Feature Spac Audio-Visual Speech R
In this paper we investigate feature space transforms to improve lip-reading performance for multi-stream HMM based audio-visual speech recognition (AVSR). The feature space transforms include non-linear Gaussianization transform and feature space maximum likelihood linear regression (fMLLR). We apply Gaussianization at the various stages of visual front-end. The results show that Gaussianizing...
متن کاملDesigning and implementing a system for Automatic recognition of Persian letters by Lip-reading using image processing methods
For many years, speech has been the most natural and efficient means of information exchange for human beings. With the advancement of technology and the prevalence of computer usage, the design and production of speech recognition systems have been considered by researchers. Among this, lip-reading techniques encountered with many challenges for speech recognition, that one of the challenges b...
متن کاملAn Efficient Lip-reading Method Using K-nearest Neighbor Algorithm
Many studies have been carried out on lip reading, most of those works are based on color images, while some essential features might not be obtained, like inner lip information. In this paper, RGBD camera will be introduced for improving the recognition rate of lip reading. We try to complete lip reading through using only gray-scale images. Thirteen groups of words are given, and we present e...
متن کاملSurface Learning with Applications to Lipreading Surface Learning with Applications to Lipreading
Most connectionist research has focused on learning mappings from one space to another (eg. classiication and regression). This paper introduces the more general task of learning constraint surfaces. It describes a simple but powerful architecture for learning and manipulating nonlinear surfaces from data. We demonstrate the technique on low dimensional synthetic surfaces and compare it to near...
متن کامل