We propose a factorized WaveNet for voice conversion with limited adaptation data. Singular value decomposition technique is applied on the dilated convolution ...

In this paper, we propose a factorized WaveNet for limited data tasks. Specifically, we apply singular value decomposition (SVD) on the dilated convolution ...

Jun 1, 2021 · In this paper, we propose a factorized WaveNet for limited data tasks. Specifically, we apply singular value decomposition (SVD) on the dilated ...

WaveNet is introduced for waveform generation. It produces high quality text-to-speech synthesis, music generation, and voice conversion.

Jun 1, 2021 · NANSY—a fully self-supervised learning system that uses perturbations to extract speech features—is proposed, which solves the problem of ...

Conference PaperPDF Available. Effective Wavenet Adaptation for Voice Conversion with Limited Data ... factorized SD. WaveNet vocoder maintained similar ...

Experimental results show that the WaveNet vocoders built using the proposed method outperform conventional STRAIGHT vocoder, and the system achieves an ...

We propose to use singular value decomposition (SVD) to reduce WaveNet parameters while maintaining its output voice quality. Specifically, we apply SVD on ...

This paper presents a statistical voice conversion (VC) tech- nique with the WaveNet-based waveform generation. VC based on a Gaussian mixture model (GMM) ...

Abstract—This paper presents a method of sequence-to- sequence (seq2seq) voice conversion using non-parallel train- ing data.