# NVlabs/noise2noise

Noise2Noise: Learning Image Restoration without Clean Data - Official TensorFlow implementation of the ICML 2018 paper

Repository: https://github.com/NVlabs/noise2noise
Canonical: https://ross.abutalabs.com/products/noise2noise
Language: Python
License: NOASSERTION
License Family: other
Last push: 2021-11-12T08:54:23+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2891, "days_push": 1755, "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 1563, forks 330 (observed 2026-08-28T04:05:04.094170+00:00)

## What it is
Official TensorFlow implementation of the Noise2Noise ICML 2018 paper, which trains image restoration networks using only corrupted (noisy) images without clean ground truth. It supports photographic denoising, Monte Carlo render denoising, and undersampled MRI reconstruction.

## Use cases
- denoise photos without clean training data
- remove noise from images using deep learning
- train an image restoration model from noisy pairs only
- denoise synthetic Monte Carlo rendered images
- reconstruct undersampled MRI scans
- reproduce the Noise2Noise paper results

## When to choose
- you want to reproduce or build on the Noise2Noise paper
- you have pairs of independently corrupted images but no clean ground truth
- you need a research baseline for blind image denoising

## When to avoid
- you need a production-ready, actively maintained denoising library
- you use modern frameworks like PyTorch
- you need a commercial license (code is CC BY-NC 4.0)
- you need up-to-date TensorFlow compatibility

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, image-processing, deep-learning
- domain: machine-learning, computer-vision, image-processing
- platform: python
- tags: denoising, tensorflow, research-code, image-restoration, mri, icml-2018, gpu, linux

## Member repositories
- NVlabs/noise2noise (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:04.094170+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 2026-08-30T04:00:14.977625+00:00, confidence not recorded.
  - readme: https://github.com/NVlabs/noise2noise (fetched 2026-08-28T04:05:04.094170+00:00, sha dfe446a05800)
- Data as of 2026-08-30T08:39:29.467469+00:00.