The second part is Facial Tracking. Facial Tracking can extract the feature points of a human subject in the synthetic speech video.

• 第二个部份是脸部追踪。脸部追踪可以从合成的语音影片中找出在人脸上的特徵点所在位置。