Multiscale voice morphing using radial basis function analysis
A new multiscale voice morphing algorithm using radial basis function (RBF) analysis is presented in this paper. The approach copes well with small training sets of high dimension, which is a problem often encountered in voice morphing. The aim of this algorithm is to transform one person's spe...
Main Authors: | , , |
---|---|
Format: | Conference item |
Published: |
2007
|
_version_ | 1797092556028248064 |
---|---|
author | Orphanidou, C Moroz, I Roberts, S |
author_facet | Orphanidou, C Moroz, I Roberts, S |
author_sort | Orphanidou, C |
collection | OXFORD |
description | A new multiscale voice morphing algorithm using radial basis function (RBF) analysis is presented in this paper. The approach copes well with small training sets of high dimension, which is a problem often encountered in voice morphing. The aim of this algorithm is to transform one person's speech pattern so that it is perceived as if it was spoken by another speaker. The voice morphing system we propose assumes parallel training data from source and target speakers and uses the theory of wavelets in order to extract speaker feature information. The spectral conversion is modelled using RBF analysis. Independent listener tests demonstrate effective transformation of the perceived speaker identity. |
first_indexed | 2024-03-07T03:47:38Z |
format | Conference item |
id | oxford-uuid:c00dacd0-abf0-4c79-85a5-a324b1ad8079 |
institution | University of Oxford |
last_indexed | 2024-03-07T03:47:38Z |
publishDate | 2007 |
record_format | dspace |
spelling | oxford-uuid:c00dacd0-abf0-4c79-85a5-a324b1ad80792022-03-27T05:51:59ZMultiscale voice morphing using radial basis function analysisConference itemhttp://purl.org/coar/resource_type/c_5794uuid:c00dacd0-abf0-4c79-85a5-a324b1ad8079Symplectic Elements at Oxford2007Orphanidou, CMoroz, IRoberts, SA new multiscale voice morphing algorithm using radial basis function (RBF) analysis is presented in this paper. The approach copes well with small training sets of high dimension, which is a problem often encountered in voice morphing. The aim of this algorithm is to transform one person's speech pattern so that it is perceived as if it was spoken by another speaker. The voice morphing system we propose assumes parallel training data from source and target speakers and uses the theory of wavelets in order to extract speaker feature information. The spectral conversion is modelled using RBF analysis. Independent listener tests demonstrate effective transformation of the perceived speaker identity. |
spellingShingle | Orphanidou, C Moroz, I Roberts, S Multiscale voice morphing using radial basis function analysis |
title | Multiscale voice morphing using radial basis function analysis |
title_full | Multiscale voice morphing using radial basis function analysis |
title_fullStr | Multiscale voice morphing using radial basis function analysis |
title_full_unstemmed | Multiscale voice morphing using radial basis function analysis |
title_short | Multiscale voice morphing using radial basis function analysis |
title_sort | multiscale voice morphing using radial basis function analysis |
work_keys_str_mv | AT orphanidouc multiscalevoicemorphingusingradialbasisfunctionanalysis AT morozi multiscalevoicemorphingusingradialbasisfunctionanalysis AT robertss multiscalevoicemorphingusingradialbasisfunctionanalysis |