Facial expression recognition based on PCA and NMF
Authors:
- 1 January 2008
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 6826-6829
Abstract
Principal Component Analysis (PCA) is a widely used technology about dimensional reduction. Non-negative Matrix Factorization (NMF), proposed by Lee and Sung, is a new image analysis method. In this paper, PCA and NMF are used to extract facial expression feature, and the recognition results of two methods are compared. We also try to process basic image matrix and weight matrix of PCA and make them as initialization of NMF. The experiments demonstrate that the method, based on the combination of PCA and NMF, has got a better recognition rate than PCA and NMF. The best recognition rate is 93.72%.Keywords
This publication has 4 references indexed in Scilit:
- Improving non-negative matrix factorizations through structured initializationPattern Recognition, 2004
- Face Recognition Using Wavelet Transform and Non-negative Matrix FactorizationLecture Notes in Computer Science, 2004
- Real time facial expression recognition with AdaBoost2004
- Non-negative Matrix Factorization for Face Recognition2002