Improving segmental GMM based voice conversion method with target frame selection | IEEE Conference Publication | IEEE Xplore