Multi-speaker Recognition System Based on Fast ICA
Abstract：It is difficult to recognize speakers when the sample voices are mixed.The paper proposes fast independent component analysis（Fast ICA） to separate individual speaker＇s voice signal from the mixed voice.Besides,a model of RBF neural network is applied to recognize the speaker.Different voice signals maintain relatively independent.Fast ICA method can be used for signal separation based on the idea of blind signal separation.As a result,features of individual speaker can be extracted from independent voice data.The model of RBF neural network achieves recognition of speakers.Experiments show that the system is able to identify individual speaker from mixed voice data.
Keywords：multi-speaker recognitionFast ICARBF neural network
Publication Date：2011-01-01
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：4( 10-13 )