{"id":15468918,"url":"https://github.com/tvercaut/densecrf","last_synced_at":"2026-05-06T12:44:21.819Z","repository":{"id":179046454,"uuid":"662766372","full_name":"tvercaut/densecrf","owner":"tvercaut","description":"GitHub archive of Philipp Krähenbühl's dense CRF code","archived":false,"fork":false,"pushed_at":"2023-07-06T10:04:32.000Z","size":1598,"stargazers_count":2,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-01-25T18:32:23.531Z","etag":null,"topics":["conditional-random-fields","crf","densecrf"],"latest_commit_sha":null,"homepage":"http://graphics.stanford.edu/projects/drf/","language":"C++","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"bsd-3-clause","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/tvercaut.png","metadata":{"files":{"readme":"README.txt","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,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2023-07-05T21:01:46.000Z","updated_at":"2023-11-08T11:20:14.000Z","dependencies_parsed_at":"2023-07-15T19:47:07.206Z","dependency_job_id":null,"html_url":"https://github.com/tvercaut/densecrf","commit_stats":{"total_commits":8,"total_committers":2,"mean_commits":4.0,"dds":0.125,"last_synced_commit":"518a20e05fb50cd33f361e72807c6b90ea3e6c91"},"previous_names":["tvercaut/densecrf"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/tvercaut%2Fdensecrf","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/tvercaut%2Fdensecrf/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/tvercaut%2Fdensecrf/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/tvercaut%2Fdensecrf/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/tvercaut","download_url":"https://codeload.github.com/tvercaut/densecrf/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":244693197,"owners_count":20494443,"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":["conditional-random-fields","crf","densecrf"],"created_at":"2024-10-02T01:46:40.748Z","updated_at":"2026-05-06T12:44:21.754Z","avatar_url":"https://github.com/tvercaut.png","language":"C++","funding_links":[],"categories":[],"sub_categories":[],"readme":"DenseCRF - Code\n=============\nhttp://graphics.stanford.edu/projects/drf/\n\nThis software pertains to the research described in the ICML 2013 paper:\nParameter Learning and Convergent Inference for Dense Random Fields, by\nPhilipp Krähenbühl and Vladlen Koltun\nand the NIPS 2011 paper:\nEfficient Inference in Fully Connected CRFs with Gaussian Edge Potentials, by\nPhilipp Krähenbühl and Vladlen Koltun\n\nIf you're using this code in a publication, please cite our papers.\n\nThis software is provided for research purposes, with absolutely no warranty\nor suggested support, and use of it most follow the BSD license agreement, at\nthe top of each source file. *Please do not contact the authors for assistance\nwith installing, understanding or running the code.* However if you think you\nhave found an interesting bug, the authors would be grateful if you could pass\non the information.\n\nChanges to the original code\n----------------------------\nThe only major difference in this released version of the code is, that I directly\ncompute the gradient of the permutohedral lattice, instead of the general Gauss\nTransform (3 line formula in p.6 in ICML 2013 paper). The gradient of the\npermutohedral lattice evaluated the exact gradient of the approximate filter.\nIn higher dimensions (\u003e3) the filter can be non continuous, which can complicate\nthe optimization. The kernel gradient is also scaled lower than other parameters,\nwhich complicates the optimization. \n\n\nHow to compile the code\n-----------------------\nDependencies:\n * cmake  http://www.cmake.org/\n * Eigen (included)\n * liblbfgs (included)\n\nLinux, Mac OS X and Windows (cygwin):\n mkdir build\n cd build\n cmake -D CMAKE_BUILD_TYPE=Release ..\n make\n cd ..\n\nWindows\n You're probably better off just copying all files into a Visual Studio\n project\n\n\nHow to run the example\n----------------------\nAn example on how to use the DenseCRF can be found in\nexamples/dense_inference.cpp. The example loads an image and some annotations.\nIt then uses a very simple classifier to compute a unary term based on those\nannotations. A dense CRF with both color dependent and color independent terms\nfind the final accurate labeling.\n\nLinux, Mac OS X and Windows (cygwin):\n build/examples/dense_inference input_image.ppm annotations.ppm output.ppm\n\nFor example:\n build/examples/dense_inference examples/im1.ppm examples/anno1.ppm output1.ppm\n\n\nAn example on how to unse the learning code can be found in \nexamples/dense_learning.cpp. The example loads a color image and ground truth\nannotation. It then learn a CRF model with a logistic regression, a label comp\nand Gaussian kernel.\n\nLinux, Mac OS X and Windows (cygwin):\n build/examples/dense_learning input_image.ppm annotations.ppm output.ppm\n\nFor example:\n build/examples/dense_learning examples/im1.ppm examples/anno1.ppm output1.ppm\n\n\nPlease note that this implementation is slightly slower than the one used to\nin our NIPS 2011 paper. Mainly because I tried to keep the code clean and easy\nto understand.\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftvercaut%2Fdensecrf","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ftvercaut%2Fdensecrf","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftvercaut%2Fdensecrf/lists"}