The lack of a publicly-available large-scale and diverse dataset has long been a significant bottleneck for singing voice applications like Singing Voice Synthesis (SVS) and Singing Voice Conversion (SVC). To tackle this problem, we present SingNet, an extensive, diverse, and in-the-wild singing voice dataset. Specifically, we propose a data processing pipeline to extract ready-to-use training data from sample packs and songs on the internet, forming 3000 hours of singing voices in various languages and styles. Furthermore, to facilitate the use and demonstrate the effectiveness of SingNet, we pre-train and open-source various state-of-the-art (SOTA) models on Wav2vec2, BigVGAN, and NSF-HiFiGAN based on our collected singing voice data. We also conduct benchmark experiments on Automatic Lyric Transcription (ALT), Neural Vocoder, and Singing Voice Conversion (SVC). Audio demos are available at:this https URL.
View on arXiv@article{gu2025_2505.09325,
title={ SingNet: Towards a Large-Scale, Diverse, and In-the-Wild Singing Voice Dataset },
author={ Yicheng Gu and Chaoren Wang and Junan Zhang and Xueyao Zhang and Zihao Fang and Haorui He and Zhizheng Wu },
journal={arXiv preprint arXiv:2505.09325},
year={ 2025 }
}