# HuCaoFighting/Swin-Unet

[ECCVW 2022] The codes for the work "Swin-Unet: Unet-like Pure Transformer for Medical Image Segmentation"

Repository: https://github.com/HuCaoFighting/Swin-Unet
Canonical: https://ross.abutalabs.com/products/swin-unet
Language: Python
License Family: other
Last push: 2025-05-09T08:44:36+00:00

## Health v2 (maintenance only)
Score: 41/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 20, release rhythm 35, longevity 100
- inputs: {"age_days": 1948, "days_push": 481, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2416, forks 367 (observed 2026-08-28T04:06:50.313135+00:00)

## What it is
Official PyTorch implementation of Swin-Unet, a U-shaped pure Transformer model for medical image segmentation, published at ECCV 2022 Medical Computer Vision Workshop. It includes training and testing scripts for the Synapse and ACDC datasets using pretrained Swin Transformer weights.

## Use cases
- segment medical images with a pure transformer model
- reproduce Swin-Unet paper results on Synapse and ACDC
- train a UNet-style transformer for organ segmentation
- compare transformer-based segmentation against CNN baselines
- fine-tune Swin Transformer for 2D medical segmentation
- use pretrained Swin weights for downstream segmentation tasks

## When to choose
- you need a pure transformer architecture for 2D medical image segmentation
- you want to reproduce or build on the ECCV 2022 Swin-Unet paper
- you have GPU resources and pretrained Swin-T weights available
- you are researching transformer-based segmentation on Synapse or ACDC datasets

## When to avoid
- you need a production-ready, licensed, maintained library (no license is provided)
- you need 3D volumetric segmentation out of the box
- you lack GPU memory or pretrained weights, which the model depends on
- you need guaranteed reproducibility across different GPU types

## Facets
- artifact type: library
- maturity: stable
- function: machine-learning, deep-learning, image-processing, computer-vision
- domain: deep-learning, computer-vision, healthcare
- platform: python
- tags: segmentation, transformer, swin-transformer, unet, medical-image-segmentation, research-code, eccv-2022, medical-imaging, gpu, linux

## Member repositories
- HuCaoFighting/Swin-Unet (main) score 41

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:50.313135+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 2026-08-30T02:33:08.666401+00:00, confidence not recorded.
  - readme: https://github.com/HuCaoFighting/Swin-Unet (fetched 2026-08-28T04:06:50.313135+00:00, sha 3f9bfaf449ce)
- Data as of 2026-08-30T08:39:29.467469+00:00.