Independent Component Analysis and Its Applications in Speech Processing
Cao Jian · Journal of Information Engineering University · 2003
Independent component analysis is a new method of signal separating technologies developed in recent years.Different from PCA and SVD,ICA's principle of separating signal,the separated components are required to be indepndent from each other,while PCA and SVD focus on the energy of the components.This paper introduces the motivation,definition and algorithm of ICA.At the end of this paper, the authors describe some experiments on the application in the field of speech processing,and the results of the experiments are given.