kevdale013's channel

We present a method for replacing facial performances in video. Our approach accounts for differences in identity, visual appearance, speech, and ...