DocumentCode :
1954215
Title :
A 2D Discrete Wavelet Transform Based 7-State Hidden Markov Model for Efficient Face Recognition
Author :
Srinivasan, M. ; Ravichandran, Naveen
Author_Institution :
Dept. of Electron. & Commun. Eng., Alpha Coll. of Eng., Chennai, India
fYear :
2013
fDate :
29-31 Jan. 2013
Firstpage :
199
Lastpage :
203
Abstract :
A novel Discrete Wavelet Transform (DWT) based on 7-States of Hidden Markov Model (HMM) for Face Recognition (FR) is proposed in this paper. To improve the accuracy of HMM based face recognition algorithm, DWT is used to replace Discrete Cosine Transform (DCT) for observation sequence vectors extraction. Extensive experiments have been conducted in our database and the FERET database shows that the proposed method can improve the accuracy significantly, especially when the face database is large and only few training images are available. As a novel point despite of five-state HMM used in pervious researches, we propose to use 7-state HMM to cover more specific details. A pre-processing procedure is introduced to reduce the complexity of the proposed system. It is evident from the outcome of these experiments that more information during training yield better results.
Keywords :
discrete wavelet transforms; face recognition; feature extraction; hidden Markov models; image sequences; vectors; 2D discrete wavelet transform; 7-state hidden Markov model; DWT; FERET face database; FR; HMM-based face recognition algorithm accuracy improvement; complexity reduction; image preprocessing procedure; observation sequence vectors extraction; training images; Databases; Discrete wavelet transforms; Face; Face recognition; Feature extraction; Hidden Markov models; Vectors; Discrete Wavelet Transform (DWT); Face Recognition (FR); Hidden Markov Model (HMM);
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Intelligent Systems Modelling & Simulation (ISMS), 2013 4th International Conference on
Conference_Location :
Bangkok
ISSN :
2166-0662
Print_ISBN :
978-1-4673-5653-4
Type :
conf
DOI :
10.1109/ISMS.2013.16
Filename :
6498264
Link To Document :
بازگشت