This repository serves as a comprehensive collection of research papers, datasets, and resources related to domain generalization (DG) in machine learning, particularly computer vision. It aims to provide a centralized hub for researchers and practitioners exploring methods to build models that perform well on unseen data distributions.

How It Works

The repository categorizes research papers by publication year and venue (e.g., CVPR, ECCV, NeurIPS, ICML), offering a structured overview of the field's progression. It also lists and describes popular datasets used for evaluating DG methods, such as Office+Caltech, PACS, and ImageNet-C, along with links to download them. The content is curated to cover various algorithmic approaches, including kernel-based, deep neural network-based, and autoencoder-based methods.

Quick Start & Requirements

This repository is a curated list of papers and datasets, not a runnable codebase. No installation or specific requirements are needed to browse its content. Links to official dataset pages and research paper code are provided where available.

Highlighted Details

• Extensive coverage of domain generalization research from 2011 to 2022.
• Detailed descriptions and download links for key DG datasets like Office+Caltech, PACS, and ImageNet-C.
• Categorization of papers by methodology (e.g., kernel-based, DNN-based, autoencoder-based).
• Includes links to code repositories for many of the cited research papers.

Maintenance & Community

The repository is maintained by amber0309. Contact information for Shoubo Hu is provided for inquiries. A list of contributors is also mentioned.

Licensing & Compatibility

The repository itself is licensed under the MIT License. The licenses of the individual datasets and research papers are independent and should be checked separately.

Limitations & Caveats

This repository is a curated list and does not provide a unified framework or codebase for domain generalization. Users will need to independently implement or find implementations for the listed research papers and datasets.