swinir 话题在 GitHub 上没有填写简介，这里按 Star 数汇总最相关的热门仓库。
Open Source Image and Video Restoration Toolbox for Super-resolution, Denoise, Deblurring, etc. Currently, it includes EDSR, RCAN, SRResNet, SRGAN, ESRGAN, EDVR, BasicVSR, SwinIR, ECBSR, etc. Also support StyleGAN2, DFDNet.
图像恢复收件箱（PyTorch）。DPIR、USRNet、DnCNN、FFDNet、SRMD、DPSR、BSRGAN、SwinIR的培训和测试代码
🌊 A Human-in-the-Loop workflow for creating HD images from text
使用NVIDIA的TensorRT和腾讯的NCNN推断进行视频帧插值和超分辨率，精心制作并打包到单个应用程序中
[ECCV] Swin2SR: SwinV2 Transformer for Compressed Image Super-Resolution and Restoration. Advances in Image Manipulation (AIM) workshop ECCV 2022. Try it out! over 3.3M runs https://replicate.com/mv-lab/swin2sr
我制作的 VitePress 网站，用于直观比较我用 600 多个不同 AI 超分辨率模型生成的输出。运行于 https://phhofm.github.io/upscale/
A High Quality (HD / 2K / 4K) Image Generation Using Stable Diffusion and Real-ESR / SwinIR /GFPGAN
A repository of models trained by me!
该项目在 GitHub 上没有填写描述
SwinIR function for VapourSynth
WindSR Dataset contains more than 22,000 pairs of HR/LR wind speed images, which are processed using the NASA's GEOS-5 Nature Run dataset. This dataset is useful for studying super-resolution for data collected using satellites rather natural RGB images.
A multistage text to image framework. Built from a inference-reduced set of min-dalle, glid-3-xl, and SwinIR.
PyTorch implementation of our paper Road Sign Classification with Denoising Pipeline Approach
Analysis resources for the Inter IIT Tech Meet 11.0 problem statement by ISRO: The Chandrayaan Moon Mapping Challenge
Transformer-based super-resolution for AMSR-2 thermal satellite imagery using SwinIR + Real-ESRGAN. Achieves 2× native upsampling with cascaded 8× capability and physics-aware loss functions.
ImageEditor is a PyQt5-based GUI application for advanced AI-powered image processing. It supports operations such as denoising, super-resolution, background removal, canvas resizing, and format conversion. It integrates powerful models such as SCUNet, SwinIR, and U²-Net.
Batch-upscale game screenshots & illustrations to 4K wallpapers using fidelity-first transformer super-resolution (ATD / SwinIR) — low AI-look, alpha-aware PNG, aspect-preserving, VRAM-tiled for consumer GPUs.
Production-grade Single Image Super-Resolution framework in pure PyTorch
SwinIR (Shifted windows Image Restoration) model for SISR (Single Image Super-Resolution) task using PyTorch
本地图片放大与 AI 修复工具，支持 Real-ESRGAN、SwinIR、批量处理和前后对比。