This dataset was recorded in quiet office and home environments, with 20 speakers participating, including 11 males and 9 females. All speakers involved in the recording were professionally selected to ensure standardized pronunciation and clear articulation. The recorded texts cover songs and other information.