This dataset was recorded in a vehicular noise environment, with 140 speakers participating, including 65 males and 75 females. All speakers who took part in the recording were professionally selected to ensure standardized pronunciation and clear enunciation. The recorded texts cover information such as numbers, dates, times, and personal names.
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.