Title :
Lip reading from scale-space measurements
Author :
Harvey, Richard ; Matthews, Iain ; Bangham, J.A. ; Cox, Stephen
Author_Institution :
Sch. of Inf. Syst., East Anglia Univ., Norwich, UK
Abstract :
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)
Keywords :
hidden Markov models; image sequences; mathematical morphology; speech recognition; HMMs; feature vector; image sequences; lip reading; mathematical morphology transform; scale measurements; scale-space measurements; spoken word; the sieve; Active shape model; Feature extraction; Hidden Markov models; Image sequences; Lips; Machine vision; Morphology; Mouth; Parametric statistics; Speech recognition;
Conference_Titel :
Computer Vision and Pattern Recognition, 1997. Proceedings., 1997 IEEE Computer Society Conference on
Conference_Location :
San Juan
Print_ISBN :
0-8186-7822-4
DOI :
10.1109/CVPR.1997.609384
