Upload a video containing a person and an audio clip of the words you want the person to say again, and generate a new video in which the person's lip movements match the content of the uploaded audio.