{"id":23841897,"url":"https://github.com/johnpertoft/noise2noise","last_synced_at":"2026-05-20T10:01:34.867Z","repository":{"id":85388314,"uuid":"147682189","full_name":"johnPertoft/noise2noise","owner":"johnPertoft","description":"Noise2Noise implementation","archived":false,"fork":false,"pushed_at":"2020-08-26T14:00:42.000Z","size":4545,"stargazers_count":3,"open_issues_count":2,"forks_count":1,"subscribers_count":2,"default_branch":"master","last_synced_at":"2025-03-22T08:48:45.255Z","etag":null,"topics":["noise2noise","tensorflow"],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/johnPertoft.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2018-09-06T13:59:42.000Z","updated_at":"2022-12-05T15:30:01.000Z","dependencies_parsed_at":"2023-04-25T02:21:02.262Z","dependency_job_id":null,"html_url":"https://github.com/johnPertoft/noise2noise","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/johnPertoft/noise2noise","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/johnPertoft%2Fnoise2noise","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/johnPertoft%2Fnoise2noise/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/johnPertoft%2Fnoise2noise/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/johnPertoft%2Fnoise2noise/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/johnPertoft","download_url":"https://codeload.github.com/johnPertoft/noise2noise/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/johnPertoft%2Fnoise2noise/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":265049214,"owners_count":23703491,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["noise2noise","tensorflow"],"created_at":"2025-01-02T18:21:55.531Z","updated_at":"2026-05-20T10:01:29.828Z","avatar_url":"https://github.com/johnPertoft.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Noise2Noise\nTensorflow [Noise2Noise](https://arxiv.org/abs/1803.04189) implementation.\n\nNoise2Noise is a machine learning algorithm that can learn signal reconstruction from only\nnoisy examples, i.e. both inputs and targets are noisy realisations of the same image.\n\n## Prerequisites\n* Tfrecord files with jpeg encoded images under key `image/encoded` for training and evaluation.\n\n## Docker\n(requires [nvidia-docker](https://github.com/NVIDIA/nvidia-docker))\n\nBuild docker image\n```bash\n$ docker build -t n2n .\n```\n\nRun a command inside docker container\n```bash\n$ ./scripts/run-in-docker \u003ccommand\u003e\n```\n\nMount extra volumes for input or output reasons. Current directory is already shared.\n```bash\n$ VOLUMES=\"/vol1:/vol1 /vol2:/vol2\" ./scripts/run-in-docker \u003ccommand\u003e\n```\n\n## Help\n```bash\n$ python -m n2n.train --helpfull\n```\n\n## Results\nImages from left to right are input image, denoised image, and ground truth noise free image.\n\n### Additive gaussian noise\n```bash\n$ python -m n2n.train \u003crequired-args\u003e --noise additive_gaussian --loss l2\n```\n![additive-gaussian-noise](results/additive-gaussian-2.png)\n\n![additive-gaussian-noise](results/additive-gaussian-gt-loss.png) ![additive-gaussian-noise](results/additive-gaussian-loss.png)\n\n\n#### Experiment with additional adversarial loss\nTODO: Compare on similar images.\n```bash\n$ python -m n2n.train \u003crequired-args\u003e --noise additive_gaussian --loss l2 --adv_loss lsgan\n```\n![additive-gaussian-noise-adv](results/additive-gaussian-adv-loss.png)\n\n### Text overlay noise\n```bash\n$ python -m n2n.train \u003crequired-args\u003e --noise text --loss l1\n```\n![text-noise](results/text.png)\n\n![text-noise](results/text-gt-loss.png) ![text-noise](results/text-loss.png)\n\n### Impulse noise\n```bash\n$ python -m n2n.train \u003crequired-args\u003e --noise impulse --loss l0\n```\n\n### Bernoulli noise\nTODO\n\n### Poisson noise\nTODO\n\n## TODO\n* Raytracing/raycasting noise?\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjohnpertoft%2Fnoise2noise","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fjohnpertoft%2Fnoise2noise","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjohnpertoft%2Fnoise2noise/lists"}