Realistic and Efficient Face Swapping: A Unified Approach with Diffusion Models

TL;DR

This paper introduces a unified diffusion-based face swapping method that improves realism and fidelity by framing the task as a self-supervised inpainting problem, utilizing CLIP features, and enabling head and accessory swapping.

Contribution

A novel diffusion model approach for face swapping that incorporates self-supervised inpainting, CLIP feature disentanglement, and mask shuffling for a universal, head-swapping capable model.

Findings

High-fidelity, realistic face swapping demonstrated on FFHQ and CelebA datasets.

Robustness to pose variation, occlusion, and color differences.

Efficient inference with minimal artifacts.

Abstract

Despite promising progress in face swapping task, realistic swapped images remain elusive, often marred by artifacts, particularly in scenarios involving high pose variation, color differences, and occlusion. To address these issues, we propose a novel approach that better harnesses diffusion models for face-swapping by making following core contributions. (a) We propose to re-frame the face-swapping task as a self-supervised, train-time inpainting problem, enhancing the identity transfer while blending with the target image. (b) We introduce a multi-step Denoising Diffusion Implicit Model (DDIM) sampling during training, reinforcing identity and perceptual similarities. (c) Third, we introduce CLIP feature disentanglement to extract pose, expression, and lighting information from the target image, improving fidelity. (d) Further, we introduce a mask shuffling technique during inpainting training, which allows us to create a so-called universal model for swapping, with an additional feature of head swapping. Ours can swap hair and even accessories, beyond traditional face swapping. Unlike prior works reliant on multiple off-the-shelf models, ours is a relatively unified approach and so it is resilient to errors in other off-the-shelf models. Extensive experiments on FFHQ and CelebA datasets validate the efficacy and robustness of our approach, showcasing high-fidelity, realistic face-swapping with minimal inference time. Our code is available at https://github.com/Sanoojan/REFace.