This repository serves as a curated collection of research papers and resources focused on text-to-image (T2I) generation. It targets researchers, engineers, and practitioners in the AI and computer vision fields, offering a structured overview of advancements, methodologies, and tools in T2I synthesis.

How It Works

The repository organizes T2I research by key directions such as general T2I, conditional generation, personalized generation, and image editing. Papers are categorized by publication year and conference, providing a chronological and thematic map of the field. It also includes sections for relevant datasets and toolkits, facilitating a comprehensive understanding of the T2I ecosystem.

Quick Start & Requirements

This repository is a curated list of research papers and does not require installation or execution. It serves as a knowledge base. Links to official papers, code repositories, and demos are provided for each entry.

Highlighted Details

• Comprehensive coverage of major conferences (CVPR, ICCV, ECCV, NeurIPS, ICLR, etc.) and arXiv preprints.
• Detailed categorization of T2I research into sub-fields like conditional generation, personalization, and editing.
• Inclusion of popular T2I products and toolkits like Stable Diffusion WebUI, Fooocus, and ComfyUI.
• Links to official papers, code, project pages, and demos for most listed research.

Maintenance & Community

The repository is actively maintained, with recent updates reflecting new publications from major conferences and arXiv. Users are encouraged to contribute via issues and pull requests.

Licensing & Compatibility

The repository itself is licensed under an unspecified license, but it primarily links to external research papers and codebases, each with their own respective licenses. Users must adhere to the licenses of the linked resources.

Limitations & Caveats

As a curated list, the repository's depth and breadth are dependent on the maintainer's selection and update frequency. It does not provide direct access to models or code, requiring users to navigate to external links.