Samples for "Unsupervised Singing Voice Conversion"

Introduction

We present supplementary audio samples that were generated using the proposed method. These samples transfer singing voices, from NUS dataset. We present below the ground truth as well as the convert songs generated for this each singer.

NUS Dataset (Male) ground truth:


JLEE JTAN VKOW SAMF KENN ZHIY

NUS Dataset (Male) conversion (or reconstruction for the same singer):

Original singer - JLEE

Original singer - JTAN

Original singer - VKOW

Original singer - SAMF

Original singer - KENN

Original singer - ZHIY

Original singer - JLEE

Original singer - JTAN

Original singer - VKOW

Original singer - SAMF

Original singer - KENN

Original singer - ZHIY