C27. Investigation about speech conversion using different techniques

Voice conversion (VC) is a process which modifies the speech signal produced by one source speaker so that it sounds like another target speaker. In this paper we compare two techniques for voice conversion. In the first technique, a conversion function based on Gaussian mixture model (GMM) is used for transforming the spectral envelope described by line spectral frequencies (LSF) parameters and the linear predictive coefficients (LPC) residuals or Mel frequency cepstral coefficients (MFCC) parameters. The second technique uses Pitch Synchronous Overlap Add (PSOLA) and resampling. The comparison between the two techniques is based on subjective evaluation, also objective evaluation such as mean squared error (MSE) and pitch estimation.