Official website | Paper(TODO)
WeSing, a publicly available high-quality Mandarin singing corpus, is designed for singing voice synthesis (SVS) systems. This corpus consists of 100 unique Mandarin songs, which were recorded by a professional female singer. All audio files were recorded with studio-quality at a sampling rate of 44,100 Hz in a professional recording studio environment.
All singing recordings have been phonetically annotated with utterance/note/phoneme boundaries and pitch types. The final dataset contains 3,756 utterances, with a total of about 6.9 hours. The testing set consists of 5 randomly chosen songs, and baseline synthesized results are provided.
The human voice is one of the most beautiful instruments. Let’s create usable singing voice synthesis technology for humanity. Enjoy!
The database is undergoing final processing and will be coming soon. Stay tuned to this project or to the official website for the latest updates. Thank you!