This repository serves as a comprehensive, citation-ranked directory of research papers, code, and resources related to Generative Adversarial Networks (GANs). It targets researchers, engineers, and students in machine learning and deep learning, providing a structured overview of the GAN landscape since its inception.

How It Works

The repository curates a vast collection of GAN-related publications, organizing them primarily by citation count to highlight influential works. It also categorizes resources into papers, theory, presentations, courses, code repositories, frameworks, and libraries, offering a multi-faceted view of the field. The inclusion of GitHub stars for code repositories helps users identify popular and well-maintained implementations.

Quick Start & Requirements

This repository is a curated list and does not require installation or execution. It serves as a reference guide.

Highlighted Details

• Features over 200 papers with direct PDF links and citation counts, sorted by influence.
• Includes links to over 50 GitHub repositories for GAN implementations, ranked by stars.
• Provides curated lists of theory papers, presentations, courses, and relevant deep learning frameworks.
• Highlights key GAN variants and applications, such as image-to-image translation, super-resolution, and text-to-image synthesis.

Maintenance & Community

The repository is presented as an "evolving repo" with an invitation for contributions via pull requests or discussions for suggestions, missing papers, or typos.

Licensing & Compatibility

The repository itself is licensed under the MIT License. The linked papers and code repositories are subject to their respective licenses.

Limitations & Caveats

The repository is a curated list and does not provide executable code or models. While comprehensive, the "quasi-exhaustive" nature means some recent or niche works might be omitted. The ranking is primarily based on citation counts and GitHub stars, which may not always reflect the absolute state-of-the-art or practical usability.