MAGICDATA Mandarin Chinese Read Speech Corpus
755-hour Mandarin Chinese read speech dataset recorded by 1,080 speakers from various accent regions, a subset of a 10,000+ hour corpus
Duration
755 hours
Languages
1
Sample Rate
16 kHz
Published
2019-05
Description
1755 hours of read speech, primarily recorded on mobile phones
21,080 speakers from various accent regions across China
3Recorded in quiet indoor environments
4Diverse text domains: interactive Q&A, music search, social messaging, smart home control, etc.
5This dataset is a subset of a larger corpus (10,566.9 hours)
6Free for academic use
Language Details
| Language | Duration |
|---|---|
| Mandarin Chinese | 755 hours |
Publisher
Resources