This dataset was recorded in a quiet office/home environment, with a total of 141 speakers participating, including 94 males and 47 females. All speakers who participated in the recording were professionally selected to ensure standard pronunciation and clear enunciation. The recorded texts cover information from news, daily conversations, tweets, and other similar content.