TCSinger 2: Customizable Multilingual Zero-shot Singing Voice Synthesis

TL;DR

TCSinger 2 is a novel multilingual zero-shot singing voice synthesis model that improves transition smoothness, style control, and robustness without relying on extensive annotations, enabling high-quality, customizable singing synthesis.

Contribution

It introduces a multi-task model with novel modules for boundary prediction, style transfer, and contrastive learning, advancing zero-shot multilingual singing voice synthesis capabilities.

Findings

Outperforms baseline models in subjective and objective metrics

Enables smooth phoneme and note transitions without annotations

Provides effective multi-level style control via diverse prompts

Abstract

Customizable multilingual zero-shot singing voice synthesis (SVS) has various potential applications in music composition and short video dubbing. However, existing SVS models overly depend on phoneme and note boundary annotations, limiting their robustness in zero-shot scenarios and producing poor transitions between phonemes and notes. Moreover, they also lack effective multi-level style control via diverse prompts. To overcome these challenges, we introduce TCSinger 2, a multi-task multilingual zero-shot SVS model with style transfer and style control based on various prompts. TCSinger 2 mainly includes three key modules: 1) Blurred Boundary Content (BBC) Encoder, predicts duration, extends content embedding, and applies masking to the boundaries to enable smooth transitions. 2) Custom Audio Encoder, uses contrastive learning to extract aligned representations from singing, speech, and textual prompts. 3) Flow-based Custom Transformer, leverages Cus-MOE, with F0 supervision, enhancing both the synthesis quality and style modeling of the generated singing voice. Experimental results show that TCSinger 2 outperforms baseline models in both subjective and objective metrics across multiple related tasks. Singing voice samples are available at https://aaronz345.github.io/TCSinger2Demo/.