Figure 11
A diagram shows two processes labeled a and b separated by a dashed line. .

Deepfake content creation pipeline. (a) Training phase: two face auto-encoders with one identical encoder and two specific decoders are trained under the supervision of face reconstruction loss and (b) Inference phase: feed forward the source face to the encoder and employ the target person’s decoder to produce the fake face.

or Create an Account

Close Modal
Close Modal