Facke: a Survey on Generative Models for Face Swapping

TL;DR

This survey evaluates various neural generative models like CVAE, CGAN, CVAE-GAN, and diffusion models for face swapping, comparing their performance and analyzing their strengths and weaknesses.

Contribution

It provides a comprehensive comparison of mainstream face swapping models and discusses potential tricks to improve their performance.

Findings

Existing models can produce highly realistic fake faces.

Comparison reveals strengths and weaknesses of each model.

Proposed tricks offer insights for future improvements.

Abstract

In this work, we investigate into the performance of mainstream neural generative models on the very task of swapping faces. We have experimented on CVAE, CGAN, CVAE-GAN, and conditioned diffusion models. Existing finely trained models have already managed to produce fake faces (Facke) indistinguishable to the naked eye as well as achieve high objective metrics. We perform a comparison among them and analyze their pros and cons. Furthermore, we proposed some promising tricks though they do not apply to this task.