This repository is a curated collection of state-of-the-art Single Image Super-Resolution (SISR) methods, serving as a comprehensive resource for researchers and practitioners in computer vision. It aims to provide an organized overview of seminal and recent advancements in SISR, facilitating learning and development in the field.

How It Works

The repository categorizes SISR methods into distinct approaches: early learning-based, sparsity-based, self-exemplars, locally linear regression, and deep architectures. This classification allows users to trace the evolution of SISR techniques, from foundational concepts like neighbor embedding and sparse representation to modern deep learning models including CNNs, GANs, and attention-based networks. The inclusion of papers and code links for each method enables direct exploration and implementation.

Quick Start & Requirements

This repository is a collection of research papers and associated code links, not a runnable software package. To utilize any specific method, users must refer to the individual paper and its linked code repository for installation and execution instructions. Prerequisites will vary significantly depending on the chosen method, often requiring specific deep learning frameworks (e.g., PyTorch, TensorFlow, Keras), Python versions, and potentially GPU acceleration.

Highlighted Details

• Comprehensive coverage of SISR methods from early learning-based approaches to cutting-edge deep learning models.
• Links to research papers and code repositories for direct access and implementation.
• Categorization of methods aids in understanding the historical development and different paradigms in SISR.
• Includes seminal works like "Learning low-level vision" and foundational deep learning papers like "Learning a deep convolutional network for image super-resolution."

Maintenance & Community

The repository is maintained by Yapeng Tian, Yunlun Zhang, and Xiaoyu Xiang. Contact information is provided for suggestions. The project is inspired by "Awesome-deep-vision" and "Awesome Computer Vision," indicating a connection to broader computer vision communities.

Licensing & Compatibility

The repository itself does not specify a license, as it is a collection of links to external research papers and code. The licensing of individual code repositories linked within the collection will vary and must be checked separately. Compatibility for commercial use or closed-source linking depends entirely on the licenses of the linked projects.

Limitations & Caveats

This is a curated list of research papers and code, not a unified framework. Users must individually assess the quality, maturity, and licensing of each linked project. There is no single installation or execution method; each linked resource requires separate setup and evaluation.