This dataset was recorded in a quiet office/home environment, with the participation of 22 speakers, including 11 males and 11 females. All speakers involved in the recording were professionally selected to ensure standard pronunciation and clear articulation, and they were scored on a word-by-word basis by linguists.