PiiGAN: Generative Adversarial Networks for Pluralistic Image Inpainting

TL;DR

PiiGAN introduces a novel style-based generative adversarial network for pluralistic image inpainting, producing diverse, high-quality results by extracting style features and guiding the generation process.

Contribution

The paper presents a new style extractor and a consistency loss in a GAN framework to generate multiple plausible inpainting results, addressing the limitation of single-output methods.

Findings

Outperforms state-of-the-art inpainting methods in quality and diversity.

Effective on datasets CelebA, PlantVillage, and MauFlex.

Generates multiple consistent inpainting results for complex images.

Abstract

The latest methods based on deep learning have achieved amazing results regarding the complex work of inpainting large missing areas in an image. But this type of method generally attempts to generate one single "optimal" result, ignoring many other plausible results. Considering the uncertainty of the inpainting task, one sole result can hardly be regarded as a desired regeneration of the missing area. In view of this weakness, which is related to the design of the previous algorithms, we propose a novel deep generative model equipped with a brand new style extractor which can extract the style feature (latent vector) from the ground truth. Once obtained, the extracted style feature and the ground truth are both input into the generator. We also craft a consistency loss that guides the generated image to approximate the ground truth. After iterations, our generator is able to learn the mapping of styles corresponding to multiple sets of vectors. The proposed model can generate a large number of results consistent with the context semantics of the image. Moreover, we evaluated the effectiveness of our model on three datasets, i.e., CelebA, PlantVillage, and MauFlex. Compared to state-of-the-art inpainting methods, this model is able to offer desirable inpainting results with both better quality and higher diversity. The code and model will be made available on https://github.com/vivitsai/PiiGAN.