Facial Expression Recognition Based on Incremental Isomap with Expression Weighted Distance

The Isometric mapping algorithm is an unsupervised manifold learning algorithm, with no consideration of the class of training samples, while supervised isometric mapping treats the difference among classes equally. Considering the inner relationship between different expressions, we have proposed isometric mapping algorithm based on expression weighted distance, which assigns weighted values according to different sample distance in order to make full use of knowledge of expression classes when calculating the geodesic distance between training samples. We use incremental isometric mapping algorithm on new samples so as to simplify computation significantly when dealing with new samples. Then k-NN classifier is applied to classify different expression features. The facial expression recognition experiments are performed on the JAFFE database and the results show that this proposed algorithm performs better than ISOMAP algorithm and supervised ISOMAP algorithm, and it is more feasible and effective. /mtp_HPc (Envelope Mean Variance method), M2M4 (Second order and Fourth order Moment method)). Furthermore, according to the characteristics of FM modulation system, we put forward the improvement of M2M4 and also discuss the optimal algorithm for underwater acoustic Rayleigh channel, Rician channel and Bellhop channel by comparing these methods.

[1]  Li Yang,et al.  Incremental Isometric Embedding of High-Dimensional Data Using Connected Neighborhood Graphs , 2009, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  Qijun Zhao,et al.  Facial expression recognition on multiple manifolds , 2011, Pattern Recognit..

[3]  Nicu Sebe,et al.  Authentic facial expression analysis , 2004, Sixth IEEE International Conference on Automatic Face and Gesture Recognition, 2004. Proceedings..

[4]  Zhang Youwei,et al.  The contrast analysis of facial expression recognition by human and computer , 2006, 2006 8th international Conference on Signal Processing.

[5]  Sun Heng-yi.  Aircraft Identification Algorithm Based on Improved ISOMAP , 2011 .

[6]  Mukund Balasubramanian,et al.  The Isomap Algorithm and Topological Stability , 2002, Science.

[7]  Daijin Kim,et al.  Natural facial expression recognition using differential-AAM and manifold learning , 2009, Pattern Recognit..

[8]  Wang Jue.  Locally Linear Embedding and Its Application in Facial Expressions Recognition , 2010 .

[9]  Flora S. Tsai.  A visualization metric for dimensionality reduction , 2012, Expert Syst. Appl..

[10]  Feng Hai.  Face Recognition Based on Gabor Wavelet and S-ISOMAP , 2008 .

[11]  Xin Geng,et al.  Supervised nonlinear dimensionality reduction for visualization and classification , 2005, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[12]  YU Xue-lian.  Radar Target Recognition Using Range Profiles Based on KLLE and KNR , 2008 .

[13]  Ziqiang Wang,et al.  Manifold Adaptive Kernel Local Fisher Discriminant Analysis for Face Recognition , 2012, J. Multim..

[14]  Ying Guang Yang,et al.  Study of Emotion Recognition Based on Surface Electromyography and Improved Least Squares Support Vector Machine , 2011, J. Comput..

[15]  Pan Chen.  Face recognition based on supervised incremental isometric mapping , 2010 .

[16]  Anil K. Jain,et al.  Nonlinear Manifold Learning for Data Stream , 2004, SDM.

[17]  MA Jian-feng.  Improved isometric mapping algorithm for manifold learning based on optimal neighborhood graph , 2011 .

[18]  J. Tenenbaum,et al.  A global geometric framework for nonlinear dimensionality reduction. , 2000, Science.

[19]  Wang Yi-jun.  Face and Expression Recognition Based on Supervised Isomap , 2009 .