Deepfakes are simple to make. A simple overview of the artificial intelligence (AI) behind deepfakes: Generative Adversarial Networks (GANs), Encoder-decoder pairs and First-Order Motion Models.

Since both encoder and decoder models are learned (meaning they can be re-trained), the same encoder or decoder architecture can be specialised for different tasks.

A person who made a translation program from 0 by deep learning will explain the mechanism in diagrams rather than complicated formulas. ... The left is the encoder and the right is the decoder.