中文标准女声音库
Open-source 10,000-sentence standard Mandarin female TTS dataset by DataBaker, ~12 hours, recorded at 48 kHz in a professional studio
Duration
12 hours
Languages
1
Sample Rate
48 kHz
Published
2019-01
Description
1Professional standard Mandarin female voice with an intellectual, warm, and natural tone, speaker aged 20-30
2Contains 10,000 utterances averaging 16 characters each, with approximately 12 hours of effective duration
3Recorded in a professional studio with SNR no less than 35 dB; recording environment and equipment remained consistent throughout
4Recording scripts cover news, fiction, technology, entertainment, dialogue, and other domains, aiming for comprehensive coverage of syllables, phonemes, tones, coarticulation, and prosody within limited data
5Annotations include phoneme-character alignment, hierarchical prosody labeling, and Chinese initial/final boundary segmentation
6Character accuracy no less than 99.8%, phoneme boundary error >10ms ratio less than 1%, syllable boundary accuracy greater than 98%
7Non-commercial use only
Language Details
| Language | Duration |
|---|---|
| Mandarin Chinese | 12 hours |
Publisher