This dataset was recorded in quiet office/home environments, with the participation of 400 speakers, including 220 males and 180 females. All speakers involved in the recording were professionally selected to ensure standardized pronunciation and clear enunciation. The recorded texts cover news and other information.