This dataset was recorded in quiet office and home environments, with 150 speakers participating, including 78 males and 72 females. All speakers involved in the recording were professionally selected to ensure standardized pronunciation and clear articulation. The recorded texts cover news, economy, entertainment, sports, and other information.