Abstract:Image processing holds an indispensable role in various facets of our daily lives, professional undertakings, and educational pursuits, encompassing a gamut of tasks including image reconstruction, inpainting, super-resolution, colorization, and editing. In recent years, the advent of advanced models rooted in Generative Adversarial Networks (GANs) has showcased remarkable capabilities in the domain of image synthesis, catapulting the direct application of these cutting-edge models to image processing to the forefront of contemporary research. Within this context, GAN inversion, an emerging paradigm, assumes a pivotal role in the landscape of image processing tasks. This paper delves into the realm of image inversion based on the latent space of GAN models. In response to the inherent limitations of current GAN inversion methods, we introduce three innovations. Firstly, we depart from the conventional use of convolutional networks for generator implementation in existing GAN inversion techniques. Our approach employs generators entirely composed of fully connected layers, marking a significant departure from spatial convolutions and information propagation across pixels. Secondly, we leverage the distinct characteristic of generators engaged in conditional independent pixel synthesis. This feature is enhanced by fusing feature maps spanning contiguous strata of a feature pyramid network during the feature extraction process. Lastly, our framework offers a high degree of versatility, extending its applicability beyond image reconstruction to domains like image inpainting, super-resolution, and image colorization. Empirical results, based on the CelebFaces Attribute-HQ (CelebA-HQ) dataset, unequivocally demonstrate that GAN inversion, built on the principle of conditional independent pixel synthesis, yields superior reconstruction outcomes. Furthermore, it proves amenable to a plethora of tasks, including image inpainting, super-resolution, and image colorization. These advances open new vistas in image processing.

GAN Inversion of High-Resolution Images

SpatialGAN: Progressive Image Generation Based on Spatial Recursive Adversarial Expansion

HRInversion: High-Resolution GAN Inversion for Cross-Domain Image Synthesis

EGAIN: Extended GAn INversion

High-Resolution GAN Inversion for Degraded Images in Large Diverse Datasets

Exploring conditional pixel-independent generation in GAN inversion for image processing

GAN Inversion: A Survey

Robust GAN inversion

In-Domain GAN Inversion for Faithful Reconstruction and Editability

High-fidelity 3D GAN Inversion by Pseudo-multi-view Optimization

Near Perfect GAN Inversion

Out-of-domain GAN Inversion Via Invertibility Decomposition for Photo-Realistic Human Face Manipulation

Force-in-domain GAN Inversion

Make Encoder Great Again in 3D GAN Inversion through Geometry and Occlusion-Aware Encoding

InvertAvatar: Incremental GAN Inversion for Generalized Head Avatars

High-Fidelity GAN Inversion with Padding Space

JoIN: Joint GANs Inversion for Intrinsic Image Decomposition

Meta-Auxiliary Network for 3D GAN Inversion

Towards Extreme Image Rescaling with Generative Prior and Invertible Prior

Towards High Resolution Video Generation with Progressive Growing of Sliced Wasserstein GANs