This dataset was recorded in quiet office and home environments, with 655 speakers participating, including 331 males and 324 females. All speakers who took part in the recording were professionally selected to ensure standardized pronunciation and clear articulation. The recorded texts cover information such as news and Twitter updates.