Soul-AILab/SoulX-Singer-Eval

A Benchmark and Evaluation Suite for Zero-shot Singing Voice Synthesis

Python34 stars3 forks

What it does

SoulX-Singer-Eval is an evaluation suite designed for zero-shot singing voice synthesis, focusing on various quality metrics such as aesthetics, pronunciation, and speaker similarity. This tool is essential for researchers and developers working on improving SVS systems, providing a standardized way to assess their performance.

Star history

Not enough history yet — 1 day(s) recorded. The daily snapshot builds this up.

Tracking

Last trending: 2026-02-21

Creator kit

Hook

Unlock the future of singing voice synthesis with SoulX-Singer-Eval – the ultimate evaluation suite for SVS systems!

Content angles

Create a tutorial on how to set up and use SoulX-Singer-Eval for evaluating SVS systems.
Discuss the importance of metrics like MOS and MCD in assessing audio quality and user experience.
Explore the datasets provided and how they can enhance the training of new SVS models.

Who should care

Researchers, developers, and enthusiasts in AI and audio synthesis looking to improve singing voice technologies.