{"id":13422664,"url":"https://github.com/satoshiiizuka/siggraph2016_colorization","last_synced_at":"2025-03-15T12:30:47.571Z","repository":{"id":47241829,"uuid":"56657547","full_name":"satoshiiizuka/siggraph2016_colorization","owner":"satoshiiizuka","description":"Code for the paper 'Let there be Color!: Joint End-to-end Learning of Global and Local Image Priors for Automatic Image Colorization with Simultaneous Classification'. ","archived":false,"fork":false,"pushed_at":"2022-06-13T02:05:02.000Z","size":1552,"stargazers_count":2158,"open_issues_count":13,"forks_count":302,"subscribers_count":112,"default_branch":"master","last_synced_at":"2024-10-27T23:59:15.520Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"http://hi.cs.waseda.ac.jp/~iizuka/projects/colorization/","language":"Lua","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/satoshiiizuka.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2016-04-20T05:20:50.000Z","updated_at":"2024-10-25T16:02:07.000Z","dependencies_parsed_at":"2022-08-29T14:10:37.439Z","dependency_job_id":null,"html_url":"https://github.com/satoshiiizuka/siggraph2016_colorization","commit_stats":null,"previous_names":[],"tags_count":1,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/satoshiiizuka%2Fsiggraph2016_colorization","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/satoshiiizuka%2Fsiggraph2016_colorization/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/satoshiiizuka%2Fsiggraph2016_colorization/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/satoshiiizuka%2Fsiggraph2016_colorization/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/satoshiiizuka","download_url":"https://codeload.github.com/satoshiiizuka/siggraph2016_colorization/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":243730999,"owners_count":20338752,"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":[],"created_at":"2024-07-30T23:00:49.983Z","updated_at":"2025-03-15T12:30:46.889Z","avatar_url":"https://github.com/satoshiiizuka.png","language":"Lua","funding_links":[],"categories":["Cool Projects","Lua","Model Zoo"],"sub_categories":["ETC"],"readme":"# [Let there be Color!: Automatic Colorization of Grayscale Images](http://hi.cs.waseda.ac.jp/~iizuka/projects/colorization/)\n\n[Satoshi Iizuka](http://hi.cs.waseda.ac.jp/~iizuka/index_eng.html)\\*, [Edgar Simo-Serra](http://hi.cs.waseda.ac.jp/~esimo/)\\*, [Hiroshi Ishikawa](http://www.f.waseda.jp/hfs/indexE.html) (\\* equal contribution)\n\n![Teaser Image](https://raw.githubusercontent.com/satoshiiizuka/siggraph2016_colorization/master/example_results.png)\n\n## Overview\n\nThis code provides an implementation of the research paper:\n\n```\n  \"Let there be Color!: Joint End-to-end Learning of Global and Local Image Priors for Automatic Image Colorization with Simultaneous Classification\"\n  Satoshi Iizuka, Edgar Simo-Serra, and Hiroshi Ishikawa\n  ACM Transaction on Graphics (Proc. of SIGGRAPH 2016), 2016\n```\n\nWe learn to automatically color grayscale images with a deep network. Our\nnetwork learns both local features and global features jointly in a single\nframework. Our approach can then be used on images of any resolution. By\nincorporating global features we are able to obtain realistic colorings with\nour model.\n\nSee our [project page](http://hi.cs.waseda.ac.jp/~iizuka/projects/colorization/) for more detailed information.\n\n## License\n\n```\n  Copyright (C) \u003c2016\u003e \u003cSatoshi Iizuka, Edgar Simo-Serra, Hiroshi Ishikawa\u003e\n\n  This work is licensed under the Creative Commons\n  Attribution-NonCommercial-ShareAlike 4.0 International License. To view a copy\n  of this license, visit http://creativecommons.org/licenses/by-nc-sa/4.0/ or\n  send a letter to Creative Commons, PO Box 1866, Mountain View, CA 94042, USA.\n\n  Satoshi Iizuka, Waseda University\n  iizuka@aoni.waseda.jp, http://hi.cs.waseda.ac.jp/~iizuka/index_eng.html\n  Edgar Simo-Serra, Waseda University\n  esimo@aoni.waseda.jp, http://hi.cs.waseda.ac.jp/~esimo/  \n```\n\n\n## Dependencies\n\n- [Torch7](http://torch.ch/docs/getting-started.html)\n- [nn](https://github.com/torch/nn)\n- [image](https://github.com/torch/image)\n- [nngraph](https://github.com/torch/nngraph)\n\nAll packages should be part of a standard Torch7 install. For information on how to install Torch7 please see the [official torch documentation](http://torch.ch/docs/getting-started.html) on the subject.\n\n## Usage\n\nFirst, download the colorization model by running the download script:\n\n```\n./download_model.sh\n```\n\nBasic usage is:\n\n```\nth colorize.lua \u003cinput_image\u003e [\u003coutput_image\u003e]\n```\n\nFor example:\n\n```\nth colorize.lua ansel_colorado_1941.png out.png\n```\n\n### Best Performance\n\n- This model was trained on the [Places dataset](http://places.csail.mit.edu/) and thus best performance is for natural outdoor images.\n- While the model works on any size image, we trained it on 224x224 pixel images and thus it works best on small images. Note that you can process a small imageto obtain the chrominance map and then rescale it and combine it with the original grayscale image for higher quality.\n- Larger image sizes can give uneven colorings (limited by spatial support of the network).\n\n### ImageNet Model\nWe also provide the colorization model that was trained on [ImageNet](http://image-net.org/challenges/LSVRC/2012/index). This model can be used for comparisons with other colorization models trained on ImageNet. We recommend using the places colorization model for general purposes.\n\nFor using the ImageNet model, download the model by running:\n\n```\n./download_model_imagenet.sh\n```\n\nUsage is:\n\n```\nth colorize.lua \u003cinput_image\u003e \u003coutput_image\u003e colornet_imagenet.t7\n```\n\n### Notes\n\n- This is developed on a linux machine running Ubuntu 14.04 during late 2015.\n- The provided code does not use GPU accelerated (trivial to change).\n- Please note that the model is slow on large images (over 512x512 pixels) and may run out of memory. Demo should take around 2 GiB of peak RAM memory, system with 4 GiB or more of RAM is recommended.\n- Provided model and sample code is under a non-commercial creative commons license.\n\n## Citing\n\nIf you use this code please cite:\n\n```\n @Article{IizukaSIGGRAPH2016,\n   author = {Satoshi Iizuka and Edgar Simo-Serra and Hiroshi Ishikawa},\n   title = {{Let there be Color!: Joint End-to-end Learning of Global and Local Image Priors for Automatic Image Colorization with Simultaneous Classification}},\n   journal = \"ACM Transactions on Graphics (Proc. of SIGGRAPH 2016)\",\n   year = 2016,\n   volume = 35,\n   number = 4,\n }\n```\n\n\n\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsatoshiiizuka%2Fsiggraph2016_colorization","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fsatoshiiizuka%2Fsiggraph2016_colorization","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsatoshiiizuka%2Fsiggraph2016_colorization/lists"}