Singaug: Data Augmentation For Singing Voice Synthesis With Cycle-consistent Training Strategy

Abstract

Deep learning based singing voice synthesis (SVS) systems have been demonstrated to flexibly generate singing with better qualities, compared to conventional statistical parametric based methods. However, neural systems are generally data-hungry and have difficulty to reach reasonable singing quality with limited public available training data. In this work, we explore different data augmentation methods to boost the training of SVS systems, including several strategies customized to SVS based on pitch augmentation and mix-up augmentation. To further stabilize the training, we introduce the cycle-consistent training strategy. Extensive experiments on two public singing databases demonstrate that our proposed augmentation methods and the stabilizing training strategy can significantly improve the performance on both objective and subjective evaluations.

Stats

• citations8
• S2 citations—
• github stars0
• HF likes0
• heat score7.16
• arxiv keyguo2022singaug

Related papers

• Visinger2+: End-to-end Singing Voice Synthesis Augmented By Self-supervised Learning Representation (2024)4.52
• Singing Voice Data Scaling-up: An Introduction To Ace-opencpop And Ace-kising (2024)15.48
• Self-supervised Singing Voice Pre-training Towards Speech-to-singing Conversion (2024)0.00
• Enhancing The Vocal Range Of Single-speaker Singing Voice Synthesis With Melody-unsupervised Pre-training (2023)3.58
• Singgan: Generative Adversarial Network For High-fidelity Singing Voice Generation (2021)10.61
• Sequence-to-sequence Singing Voice Synthesis With Perceptual Entropy Loss (2020)9.59
• Singing Voice Separation: A Study On Training Data (2019)10.07
• Improving Singing Voice Separation Using Deep U-net And Wave-u-net With Data Augmentation (2019)10.35