# swz30/Restormer

[CVPR 2022--Oral] Restormer: Efficient Transformer for High-Resolution Image Restoration. SOTA  for motion deblurring, image deraining, denoising (Gaussian/real data), and defocus deblurring.

Repository: https://github.com/swz30/Restormer
Canonical: https://ross.abutalabs.com/products/restormer
Language: Python
License: MIT
License Family: permissive
Topics: image-restoration, image-deraining, image-deblurring, defocus-deblurring, motion-deblurring, transformer, pytorch, low-level-vision, cvpr2022, high-resolution, efficient-transformers
Last push: 2025-10-23T13:41:29+00:00

## Health v2 (maintenance only)
Score: 44/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 48, release rhythm 8, longevity 100
- inputs: {"age_days": 1779, "days_push": 314, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2625, forks 329 (observed 2026-08-28T04:07:05.294113+00:00)

## What it is
Restormer is an efficient Transformer architecture for high-resolution image restoration, published as a CVPR 2022 Oral paper. It provides PyTorch training and testing code with pre-trained models for motion deblurring, deraining, denoising, and defocus deblurring.

## Use cases
- remove motion blur from photos
- denoise images with gaussian or real noise
- remove rain streaks from images
- correct defocus blur in pictures
- restore high-resolution images with a transformer model
- train an image restoration model in pytorch

## When to choose
- you need state-of-the-art restoration for deblurring, deraining, or denoising
- you work with high-resolution images where standard vision transformers are too slow
- you want pre-trained models plus training code for low-level vision research

## When to avoid
- you need a production-ready image editing application with a polished UI
- your task is high-level vision like detection or segmentation rather than restoration
- you lack a GPU and cannot afford computationally heavy inference

## Facets
- artifact type: library
- maturity: stable
- function: image-processing, machine-learning, deep-learning
- domain: computer-vision, image-processing, deep-learning, artificial-intelligence
- platform: python, cross-platform
- tags: image-restoration, image-deblurring, image-deraining, denoising, transformer, pytorch, low-level-vision, cvpr2022, high-resolution, research-code, gpu

## Member repositories
- swz30/Restormer (main) score 44

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:05.294113+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:19:51.811311+00:00, confidence not recorded.
  - readme: https://github.com/swz30/Restormer (fetched 2026-08-28T04:07:05.294113+00:00, sha 9d46f1669158)
- Data as of 2026-08-30T08:39:29.467469+00:00.