This dataset was recorded in a vehicular noise environment, with a total of 66 speakers participating, including 34 males and 32 females. All participants in the recording were professionally selected to ensure standardized pronunciation and clear articulation. The recorded texts span information from news, everyday conversations, tweets, and other relevant content.