This dataset was recorded in a vehicular noise environment, with the participation of 300 speakers, including 152 males and 148 females. All speakers involved in the recording were carefully selected by professionals to ensure standardized pronunciation and clear articulation. 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.