TokSing: Singing Voice Synthesis based on Discrete Tokens

Recent advancements in speech synthesis witness significant benefits by leveraging discrete tokens extracted from self-supervised learning (SSL) models. Discrete tokens offer higher storage efficiency and greater operability in intermediate representations compared to traditional continuous Mel spec...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2024-06
Main Authors: Wu, Yuning, zhang, Chunlei, Shi, Jiatong, Tang, Yuxun, Yang, Shan, Qin, Jin
Format: Article
Language:English
Subjects:
Representations
Self-supervised learning
Singing
Spectrograms
Speech recognition
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Recent advancements in speech synthesis witness significant benefits by leveraging discrete tokens extracted from self-supervised learning (SSL) models. Discrete tokens offer higher storage efficiency and greater operability in intermediate representations compared to traditional continuous Mel spectrograms. However, when it comes to singing voice synthesis(SVS), achieving higher levels of melody expression poses a great challenge for utilizing discrete tokens. In this paper, we introduce TokSing, a discrete-based SVS system equipped with a token formulator that offers flexible token blendings. We observe a melody degradation during discretization, prompting us to integrate a melody signal with the discrete token and incorporate a specially-designed melody enhancement strategy in the musical encoder. Extensive experiments demonstrate that our TokSing achieves better performance against the Mel spectrogram baselines while offering advantages in intermediate representation space cost and convergence speed.
ISSN:2331-8422