{"id":19543165,"url":"https://github.com/yeonghyeon/adae-tf","last_synced_at":"2026-04-16T22:31:28.395Z","repository":{"id":159175926,"uuid":"348919356","full_name":"YeongHyeon/ADAE-TF","owner":"YeongHyeon","description":"TensorFlow implementation of Anomaly Detection with Adversarial Dual Autoencoders (ADAE) with MNIST dataset.","archived":false,"fork":false,"pushed_at":"2021-03-26T09:13:07.000Z","size":1929,"stargazers_count":4,"open_issues_count":0,"forks_count":0,"subscribers_count":2,"default_branch":"master","last_synced_at":"2025-02-26T05:32:53.471Z","etag":null,"topics":["adversarial-dual-autoencoders","anomaly-detection","mnist","mnist-dataset","tensorflow"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/YeongHyeon.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,"governance":null}},"created_at":"2021-03-18T02:44:43.000Z","updated_at":"2024-08-02T00:33:20.000Z","dependencies_parsed_at":"2023-07-28T19:00:58.021Z","dependency_job_id":null,"html_url":"https://github.com/YeongHyeon/ADAE-TF","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/YeongHyeon/ADAE-TF","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/YeongHyeon%2FADAE-TF","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/YeongHyeon%2FADAE-TF/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/YeongHyeon%2FADAE-TF/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/YeongHyeon%2FADAE-TF/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/YeongHyeon","download_url":"https://codeload.github.com/YeongHyeon/ADAE-TF/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/YeongHyeon%2FADAE-TF/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31907434,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-16T18:22:33.417Z","status":"ssl_error","status_checked_at":"2026-04-16T18:21:47.142Z","response_time":69,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.6:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"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":["adversarial-dual-autoencoders","anomaly-detection","mnist","mnist-dataset","tensorflow"],"created_at":"2024-11-11T03:17:46.568Z","updated_at":"2026-04-16T22:31:28.369Z","avatar_url":"https://github.com/YeongHyeon.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"[TensorFlow] Anomaly Detection with Adversarial Dual Autoencoders\n=====\n\nTensorFlow implementation of Anomaly Detection with Adversarial Dual Autoencoders (ADAE) with MNIST dataset.  \nThe Keras implementation is provided as the following link.  \nhttps://github.com/kjm1559/ADAE_LSTM_Autoencoder\n\n## Architecture\n\n### Objective Functions\n\u003cdiv align=\"center\"\u003e\n  \u003cimg src=\"./figures/losses.png\" width=\"500\"\u003e  \n  \u003cp\u003eThe objective functions (losses) for training ADAE [1].\u003c/p\u003e\n\u003c/div\u003e\n\n### ADAE architecture\n\u003cdiv align=\"center\"\u003e\n  \u003cimg src=\"./figures/adae.png\" width=\"500\"\u003e  \n  \u003cp\u003eThe architecture of ADAE.\u003c/p\u003e\n\u003c/div\u003e\n\n### Graph in TensorBoard\n\u003cdiv align=\"center\"\u003e\n  \u003cimg src=\"./figures/graph.png\" width=\"650\"\u003e  \n  \u003cp\u003eGraph of ADAE.\u003c/p\u003e\n\u003c/div\u003e\n\n### Problem Definition\n\u003cdiv align=\"center\"\u003e\n  \u003cimg src=\"./figures/definition.png\" width=\"450\"\u003e  \n  \u003cp\u003e'Class-1' is defined as normal and the others are defined as abnormal.\u003c/p\u003e\n\u003c/div\u003e\n\n## Results\n\n### Training Procedure\n\u003cdiv align=\"center\"\u003e\n  \u003cp\u003e\n    \u003cimg src=\"./figures/ADAE_G_loss_g.svg\" width=\"250\"\u003e\n    \u003cimg src=\"./figures/ADAE_G_loss_g_term1.svg\" width=\"250\"\u003e\n    \u003cimg src=\"./figures/ADAE_G_loss_g_term2.svg\" width=\"250\"\u003e\n  \u003c/p\u003e\n  \u003cp\u003eLoss graphs in the training procedure.\u003c/br\u003eEach graph shows the generative loss, and the two terms that make loss-G.\u003c/p\u003e\n\u003c/div\u003e\n\n\u003cdiv align=\"center\"\u003e\n  \u003cp\u003e\n    \u003cimg src=\"./figures/ADAE_G_loss_g.svg\" width=\"250\"\u003e\n    \u003cimg src=\"./figures/ADAE_G_loss_g_term1.svg\" width=\"250\"\u003e\n    \u003cimg src=\"./figures/ADAE_G_loss_g_term2.svg\" width=\"250\"\u003e\n  \u003c/p\u003e\n  \u003cp\u003eLoss graphs in the training procedure.\u003c/br\u003eEach graph shows the discriminative loss, and the two terms that make loss-G.\u003c/p\u003e\n\u003c/div\u003e\n\n\u003cdiv align=\"center\"\u003e\n  \u003cimg src=\"./figures/restoring.png\" width=\"800\"\u003e  \n  \u003cp\u003eRestoration result by ADAE.\u003c/p\u003e\n\u003c/div\u003e\n\n### Test Procedure\n\u003cdiv align=\"center\"\u003e\n  \u003cimg src=\"./figures/test-box.png\" width=\"400\"\u003e\n  \u003cp\u003eBox plot with encoding loss of test procedure.\u003c/p\u003e\n\u003c/div\u003e\n\n\u003cdiv align=\"center\"\u003e\n  \u003cp\u003e\n    \u003cimg src=\"./figures/in_in01.png\" width=\"130\"\u003e\n    \u003cimg src=\"./figures/in_in02.png\" width=\"130\"\u003e\n    \u003cimg src=\"./figures/in_in03.png\" width=\"130\"\u003e\n  \u003c/p\u003e\n  \u003cp\u003eNormal samples classified as normal.\u003c/p\u003e\n\n  \u003cp\u003e\n    \u003cimg src=\"./figures/in_out01.png\" width=\"130\"\u003e\n    \u003cimg src=\"./figures/in_out02.png\" width=\"130\"\u003e\n    \u003cimg src=\"./figures/in_out03.png\" width=\"130\"\u003e\n  \u003c/p\u003e\n  \u003cp\u003eAbnormal samples classified as normal.\u003c/p\u003e\n\n  \u003cp\u003e\n    \u003cimg src=\"./figures/out_in01.png\" width=\"130\"\u003e\n    \u003cimg src=\"./figures/out_in02.png\" width=\"130\"\u003e\n    \u003cimg src=\"./figures/out_in03.png\" width=\"130\"\u003e\n  \u003c/p\u003e\n  \u003cp\u003eNormal samples classified as abnormal.\u003c/p\u003e\n\n  \u003cp\u003e\n    \u003cimg src=\"./figures/out_out01.png\" width=\"130\"\u003e\n    \u003cimg src=\"./figures/out_out02.png\" width=\"130\"\u003e\n    \u003cimg src=\"./figures/out_out03.png\" width=\"130\"\u003e\n  \u003c/p\u003e\n  \u003cp\u003eAbnormal samples classified as abnormal.\u003c/p\u003e\n\u003c/div\u003e\n\n\n## Environment\n* Python 3.7.4  \n* Tensorflow 1.14.0  \n* Numpy 1.17.1  \n* Matplotlib 3.1.1  \n* Scikit Learn (sklearn) 0.21.3  \n\n\n## Reference\n[1] Ha Son Vu et al. (2019). \u003ca href=\"https://arxiv.org/abs/1902.06924\"\u003eAnomaly Detection with Adversarial Dual Autoencoders\u003c/a\u003e. arXiv preprint arXiv:1902.06924.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fyeonghyeon%2Fadae-tf","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fyeonghyeon%2Fadae-tf","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fyeonghyeon%2Fadae-tf/lists"}