One-shot domain adaptation for semantic face editing of real world images using StyleALAE

TL;DR

This paper introduces a one-shot domain adaptation method for semantic face editing that preserves identity in real-world images using styleALAE, avoiding extensive training of encoder networks.

Contribution

It proposes a novel approach combining styleALAE with one-shot domain adaptation to enable identity-preserving semantic face editing without additional training.

Findings

Successfully preserves identity during semantic editing

Enables editing of real-world facial images

Avoids time-consuming encoder training

Abstract

Semantic face editing of real world facial images is an important application of generative models. Recently, multiple works have explored possible techniques to generate such modifications using the latent structure of pre-trained GAN models. However, such approaches often require training an encoder network and that is typically a time-consuming and resource intensive process. A possible alternative to such a GAN-based architecture can be styleALAE, a latent-space based autoencoder that can generate photo-realistic images of high quality. Unfortunately, the reconstructed image in styleALAE does not preserve the identity of the input facial image. This limits the application of styleALAE for semantic face editing of images with known identities. In our work, we use a recent advancement in one-shot domain adaptation to address this problem. Our work ensures that the identity of the reconstructed image is the same as the given input image. We further generate semantic modifications over the reconstructed image by using the latent space of the pre-trained styleALAE model. Results show that our approach can generate semantic modifications on any real world facial image while preserving the identity.