CVAIJun 22, 2022

Facke: a Survey on Generative Models for Face Swapping

arXiv:2206.11203v13 citationsh-index: 21
Originality Synthesis-oriented
AI Analysis

This work provides a comparative analysis for researchers in face swapping, but it is incremental as it surveys and evaluates existing methods without introducing a new approach.

The paper compares the performance of mainstream neural generative models, including CVAE, CGAN, CVAE-GAN, and conditioned diffusion models, on face swapping, finding that existing models can produce fake faces indistinguishable to the naked eye and achieve high objective metrics.

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.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes