This repository is a curated collection of research papers and resources focused on image inpainting, a computer vision task that involves filling in missing or corrupted parts of an image. It serves as a comprehensive reference for researchers and practitioners in the field, offering categorized links to papers, datasets, and related tools.

How It Works

The repository organizes image inpainting research chronologically and by methodology, categorizing papers into "Deep Learning-Based" (further subdivided by task, e.g., text-guided, sketch-guided, face inpainting) and "Conventional Methods." Each entry typically includes a link to the paper, and often to associated code repositories, project pages, or demos, facilitating easy access to implementations and further details.

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.

Highlighted Details

• Extensive categorization of papers by sub-task (e.g., text-guided, sketch-guided, face, fashion) and methodology.
• Chronological organization of research from 2000 to the present, highlighting the evolution of techniques.
• Includes links to relevant datasets (Places, ImageNet, CelebA-HQ, FFHQ) and survey papers for broader context.
• Provides direct links to code repositories and project pages for many papers, enabling practical exploration.

Maintenance & Community

The repository is actively maintained, with recent updates reflecting papers from late 2023 and early 2024. Suggestions and contributions are welcomed via GitHub issues and pull requests.

Licensing & Compatibility

This repository is a collection of links to external resources and does not have its own license. The licensing of individual papers and code repositories varies and must be checked on their respective sources.

Limitations & Caveats

The repository is a curated list and does not provide any executable code or models itself. Availability of linked code repositories or datasets may change over time.