Please use this identifier to cite or link to this item:
Title: Redes neurais e máquinas de vetores de suporte no reconhecimento de locutor usando coeficientes MFC e características do sinal glotal
Other Titles: Neural networks and machines of support vector machines speaker recognition user MFK and coefficients characteristics of the glotal signal
Keywords: Reconhecimento de locutor;  Filtragem inversa;  Máquina de vetores de suporte;  Codificador de voz;  Sinal glotal;  Voz;  Rede neural;  Recognition speech;  Inverse filtering;  Support vector machine
Issue Date: 21-Aug-2009
Abstract: This work presents a proposal for automatic speaker recognition using support vector machine and neural networks. The input vector is hybrid composed by MFC coefficients (Mel Frequency Cepstrum Coefficients) and features extracted from the glottal signal, obtained by inverse filtering of the voice signal. The results are compared with other obteined when only the MFC coefficients are used as input.
Appears in Collections:TEDE sem arquivo

Files in This Item:
There are no files associated with this item.

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.