Title :
Residual prediction
Author :
Sündermann, David ; Höge, Harald ; Bonafonte, Antonio ; Duxans, Helenca
Author_Institution :
Siemens AG, Munich
Abstract :
Residual prediction is a technique that aims at recovering the spectral details of speech that was encoded using parameterizations as linear predictive coefficients. Example applications of residual prediction are hidden Markov model-based speech synthesis or voice conversion. Our voice conversion experiments showed that only one of the seven compared techniques was capable of successfully converting the voice while achieving a fair speech quality (i.e. mean opinion score = 3)
Keywords :
hidden Markov models; speech coding; speech synthesis; hidden Markov model; linear predictive coefficients; residual prediction; speech synthesis; voice conversion; Frequency conversion; Frequency synthesizers; Hidden Markov models; Interpolation; Mel frequency cepstral coefficient; Predictive models; Pulse generation; Signal processing algorithms; Speech synthesis; Vectors;
Conference_Titel :
Signal Processing and Information Technology, 2005. Proceedings of the Fifth IEEE International Symposium on
Conference_Location :
Athens
Print_ISBN :
0-7803-9313-9
DOI :
10.1109/ISSPIT.2005.1577150
