Title :
Result based analysis of various lip tracking systems
Author :
Sahu, Vivek ; Sharma, Mukesh
Abstract :
With the advancement of technologies, an Automatic Lip reading system has always been requirement in various safety and security concerned systems. Moreover under noisy conditions it plays a very important role for listeners with a hearing impairment. Study of techniques associated with various Visual Speech Recognition Systems, which are developed till now, is done in this paper. Results associated with lip feature extraction techniques, employed in these systems, are analyzed. On the basis of analysis, it is found, that accuracy of Lip Geometry Estimation and Image Based Raw Information Techniques are 91% and 92-93% respectively. Hence on an average these techniques are giving best results for lip feature extraction.
Keywords :
feature extraction; geometry; handicapped aids; object tracking; speech recognition; automatic lip reading system; hearing impairment; image based raw information techniques; lip feature extraction techniques; lip geometry estimation; lip tracking system; result based analysis; visual speech recognition systems; Estimation; Feature extraction; Hidden Markov models; Mouth; Speech; Speech recognition; Visualization; AVSR; Hybrid Approach; Image Transform; Lip Geometry Estimation; Model Based Tracking; Point tracking;
Conference_Titel :
Green High Performance Computing (ICGHPC), 2013 IEEE International Conference on
Conference_Location :
Nagercoil
Print_ISBN :
978-1-4673-2592-9
DOI :
10.1109/ICGHPC.2013.6533911
