Two Birds with One Stone: Transforming and Generating Facial Images with Iterative GAN

TL;DR

This paper introduces an iterative GAN framework that combines pixel and perceptual losses to improve high-fidelity, identity-preserving facial image generation and attribute transformation, outperforming existing methods.

Contribution

It proposes a novel iterative GAN architecture with integrated loss functions for simultaneous face generation and attribute transformation, emphasizing perceptual information.

Findings

High-quality, identity-preserving facial images generated.

Effective multi-attribute recognition demonstrated.

Controllable facial attribute transformation achieved.

Abstract

Generating high fidelity identity-preserving faces with different facial attributes has a wide range of applications. Although a number of generative models have been developed to tackle this problem, there is still much room for further improvement.In paticular, the current solutions usually ignore the perceptual information of images, which we argue that it benefits the output of a high-quality image while preserving the identity information, especially in facial attributes learning area.To this end, we propose to train GAN iteratively via regularizing the min-max process with an integrated loss, which includes not only the per-pixel loss but also the perceptual loss. In contrast to the existing methods only deal with either image generation or transformation, our proposed iterative architecture can achieve both of them. Experiments on the multi-label facial dataset CelebA demonstrate that the proposed model has excellent performance on recognizing multiple attributes, generating a high-quality image, and transforming image with controllable attributes.