# zhixuhao/unet

unet for image segmentation

Repository: https://github.com/zhixuhao/unet
Canonical: https://ross.abutalabs.com/products/unet
Language: Jupyter Notebook
License: MIT
License Family: permissive
Topics: unet, keras, segmentation
Last push: 2026-03-27T03:48:07+00:00

## Health v2 (maintenance only)
Score: 66/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 74, release rhythm 35, longevity 100
- inputs: {"age_days": 3437, "days_push": 159, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4941, forks 2001 (observed 2026-08-28T04:09:03.569646+00:00)

## What it is
A Keras implementation of the U-Net convolutional network architecture for image segmentation, based on the original biomedical segmentation paper. It includes training scripts, data augmentation via Keras ImageDataGenerator, and a sample ISBI membrane dataset.

## Use cases
- segment biomedical images like cell membranes
- train a U-Net model on my own image masks
- do binary image segmentation with Keras
- learn how U-Net is implemented in Keras
- apply data augmentation to a small segmentation dataset
- predict segmentation masks for test images

## When to choose
- you want a simple, well-known Keras U-Net reference implementation
- you are doing binary segmentation on small datasets with augmentation
- you want a teaching example or starting point to modify the architecture

## When to avoid
- you need multi-class segmentation or modern architectures out of the box
- you need actively maintained code compatible with recent TensorFlow/Keras versions
- you need a production-grade segmentation pipeline with evaluation metrics

## Facets
- artifact type: library
- maturity: maintenance
- function: deep-learning, image-processing, computer-vision, machine-learning
- domain: deep-learning, computer-vision, image-processing, machine-learning
- platform: python, cross-platform
- tags: unet, keras, tensorflow, semantic-segmentation, biomedical-imaging, jupyter-notebook, gpu

## Member repositories
- zhixuhao/unet (main) score 66

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:03.569646+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 2026-08-29T18:17:55.182523+00:00, confidence not recorded.
  - readme: https://github.com/zhixuhao/unet (fetched 2026-08-28T04:09:03.569646+00:00, sha 9505474379ad)
- Data as of 2026-08-30T08:39:29.467469+00:00.