{"id":13546342,"url":"https://github.com/musyoku/ddgm","last_synced_at":"2025-04-02T18:30:35.678Z","repository":{"id":202249051,"uuid":"69355141","full_name":"musyoku/ddgm","owner":"musyoku","description":"Chainer implementation of Deep Directed Generative Models with Energy-Based Probability Estimation","archived":false,"fork":false,"pushed_at":"2016-11-20T05:59:43.000Z","size":134,"stargazers_count":2,"open_issues_count":0,"forks_count":1,"subscribers_count":3,"default_branch":"master","last_synced_at":"2024-11-03T14:35:35.592Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","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/musyoku.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}},"created_at":"2016-09-27T12:40:03.000Z","updated_at":"2020-06-22T10:36:19.000Z","dependencies_parsed_at":null,"dependency_job_id":"fd633dd4-64d1-4b64-8376-1a7d54969f8f","html_url":"https://github.com/musyoku/ddgm","commit_stats":null,"previous_names":["musyoku/ddgm"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/musyoku%2Fddgm","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/musyoku%2Fddgm/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/musyoku%2Fddgm/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/musyoku%2Fddgm/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/musyoku","download_url":"https://codeload.github.com/musyoku/ddgm/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":246869599,"owners_count":20847157,"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":"2024-08-01T12:00:35.705Z","updated_at":"2025-04-02T18:30:35.361Z","avatar_url":"https://github.com/musyoku.png","language":"Python","funding_links":[],"categories":["Preferred Networks Research"],"sub_categories":["Generative models"],"readme":"## Deep Directed Generative Models with Energy-Based Probability Estimation\n\ncode for the [paper](https://arxiv.org/abs/1606.03439)\n\n[この記事](http://musyoku.github.io/2016/10/28/Deep-Directed-Generative-Models-with-Energy-Based-Probability-Estimation/)で実装したコードです。\n\n### Requirements\n\n- Chainer 1.17\n- PIL\n- pylab\n\nContains the following repository:\n\n- [chainer-sequential](https://github.com/musyoku/chainer-sequential)\n\n## 2D datasets\n\nTrain generator to generate 10 dimensional Gaussian mixture distribution and swiss-roll distribution.\n\n![gaussian](https://github.com/musyoku/musyoku.github.io/blob/master/images/post/2016-10-28/gaussian.png?raw=true)\n![swiss_roll](https://github.com/musyoku/musyoku.github.io/blob/master/images/post/2016-10-28/swissroll.png?raw=true)\n\nSee videos:\n\n- [https://gfycat.com/DarlingShowyHypsilophodon](https://gfycat.com/DarlingShowyHypsilophodon)\n- [https://gfycat.com/UnrulyMisguidedHornedviper](https://gfycat.com/UnrulyMisguidedHornedviper)\n\n### Running\n\nrun `train_2d/train.py` to train the model.\n\nrun `train_2d/gif_gaussian.py` or `train_2d/gif_swissroll.py` to generate gif frames.\n\n## MNIST\n\nrun `train_mnist/train.py`\n\nIf there is no MNIST image, it will be downloaded automatically.\n\n### Genereted images\n\n![result](https://github.com/musyoku/musyoku.github.io/blob/master/images/post/2016-10-28/mnist_success.png?raw=true)\n\n## killmebaby（キルミーベイベー）\n\nDownload 686 images from [http://killmebaby.tv/special_icon.html](http://killmebaby.tv/special_icon.html) and resize all to 64x64 pixels.\n\nrun `train_killmebaby/train.py` \n\n### Original images\n\n![original](https://github.com/musyoku/musyoku.github.io/blob/master/images/post/2016-10-28/kb_original.png?raw=true)\n\n\n### Images generated by Deep Generative Model\n\n![gen](https://github.com/musyoku/musyoku.github.io/blob/master/images/post/2016-10-28/kb_gen.png?raw=true)\n\nSince the position of the face of the training data is not constant, I think it is difficult to train the generator, but relatively clean images are generated.\n\n### When learning of Generator did not go well\n\n![gen](https://github.com/musyoku/musyoku.github.io/blob/master/images/post/2016-10-28/kb_fail.png?raw=true)\n\nWhichever noise z is used to generate an image, an average is generated.","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmusyoku%2Fddgm","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmusyoku%2Fddgm","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmusyoku%2Fddgm/lists"}