This page presents our paper:
Title: Discrete Optimal Transport and Voice Conversion
Authors: Anton Selitskiy and Maitreya Kocharekar
PDF: Download paper
Code: on GitHub
In this work, we address the voice conversion (VC) task using a vector-based interface. To align audio embeddings between speakers, we employ discrete optimal transport mapping. Our evaluation results demonstrate the high quality and effectiveness of this method. Additionally, we show that applying discrete optimal transport as a post-processing step in audio generation can lead to the incorrect classification of synthetic audio as real.
Link to examples, or embed a player using HTML.
Source and reference are up to 1 min.
| Source | Reference |
|---|---|
|
k = 1 |
k = 3 |
k = 4 |
|
k = 5 |
k = 10 |
k = 40 |
Source is represented by 1 utterance, reference uses 10 utterances
| Source | Reference |
|---|---|
|
k = 1 |
k = 3 |
k = 4 |
|
k = 5 |
k = 10 |
k = 40 |
BibTeX entry or link to citation.
Created with ❤️ using GitHub Pages.