This dataset was recorded in quiet office and home environments, involving a total of 2,575 speakers, with 1,355 males and 1,220 females. All speakers who participated in the recording were professionally screened to ensure standard pronunciation and clear enunciation. The recorded texts cover various types of information, including everyday conversations and news updates.
Chinese English & American English Speech Recognition Corpus (desktop)
This dataset includes 6 topic for Office White-collar Meeting Scenario - IT and Internet, Finance, Clear Energy, Healthcare, Media and Consumer Electronics.