Note: This bibliographic page is archived and will no longer be updated. For an up-to-date list of publications from the Music Technology Group see the Publications list .

Observation-Model Error Compensation for Enhanced Spectral Envelope Transformation in Voice Conversion

Title Observation-Model Error Compensation for Enhanced Spectral Envelope Transformation in Voice Conversion
Publication Type Conference Paper
Year of Publication 2015
Conference Name IEEE International Workshop on Machine Learning for Signal Processing
Authors Villavicencio, F. , Bonada J. , & Hisaminato Y.
Conference Start Date 17/07/2015
Conference Location Boston, USA
Abstract This work proposes a novel derivation of the spectral envelope transformation in Voice Conversion to alleviate degradations in the converted speech quality produced by the imposition of oversmoothed spectra. The existing mismatch between an input feature and the corresponding observation by the statistical model denotes an averaging of the features due to the model’s limited capacity to represent the feature space. The proposition is based on compensating this mismatch on the transformation applied to the input spectra. As a result, the perceived naturalness of the converted speech is enhanced. Our claim is supported by the results of objective and subjective evaluations comparing speech converted by the conventional transformation and the proposed one.