{"id":15646803,"url":"https://github.com/sseung0703/sskd_svd","last_synced_at":"2025-04-30T12:29:23.923Z","repository":{"id":119339109,"uuid":"125353128","full_name":"sseung0703/SSKD_SVD","owner":"sseung0703","description":null,"archived":false,"fork":false,"pushed_at":"2019-08-08T17:54:14.000Z","size":1731,"stargazers_count":49,"open_issues_count":5,"forks_count":10,"subscribers_count":5,"default_branch":"master","last_synced_at":"2024-10-03T12:19:30.274Z","etag":null,"topics":["deep-learning","knowledge-distillation"],"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/sseung0703.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":"2018-03-15T10:49:39.000Z","updated_at":"2024-03-15T08:33:16.000Z","dependencies_parsed_at":null,"dependency_job_id":"9dcf4883-4883-4688-8858-22fed648800a","html_url":"https://github.com/sseung0703/SSKD_SVD","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sseung0703%2FSSKD_SVD","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sseung0703%2FSSKD_SVD/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sseung0703%2FSSKD_SVD/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sseung0703%2FSSKD_SVD/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/sseung0703","download_url":"https://codeload.github.com/sseung0703/SSKD_SVD/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":221113218,"owners_count":16758425,"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":["deep-learning","knowledge-distillation"],"created_at":"2024-10-03T12:15:00.518Z","updated_at":"2024-10-22T16:08:08.175Z","avatar_url":"https://github.com/sseung0703.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"More clear codes are provided at https://github.com/sseung0703/KD_methods_with_TF.\n# Self-supervised Knowledge Distillation using Singular Value Decomposition\n![Alt text](dist.png)\n## Feature\n- Define knowledge by Singular value decomposition\n- Fast and efficient learning by multi-task learning\n \n## Requirments\n- Tensorflow\n- Scipy\n\nUnfortunatly SVD is very slow on GPU. so  recommend below installation method.\n- Install Tensorflow from source which is removed SVD GPU op.(recommended)\n- Install ordinary Tensorflow and make SVD using CPU.\n- Install Tensorflow version former than 1.2.\n\n## How to Use\nThe code is based on Tensorflow-slim example codes. so if you used that it is easy to understand.\n( If you want recoded dataset and trained teacher weights download them on https://drive.google.com/open?id=1sNPkr-hOy3qKNUE8bu5qJpvCGFI7VOJN )\n1. Recording Cifar100 dataset to tfrecording file \n2. Train teacher network by train_teacher.py, and you can find trained_param.mat in training path.\n3. Train student network using teacher knowledge which trained step 2 or default knowledge which is ImageNet pretrained VGG16.\n\n## Results\n![Alt text](results.png)\n\n## Paper\nThis research accepted to ECCV2018 poster session\nand arxiv version paper is available on https://arxiv.org/abs/1807.06819\nand ECCV format paper is available on http://openaccess.thecvf.com/content_ECCV_2018/papers/SEUNG_HYUN_LEE_Self-supervised_Knowledge_Distillation_ECCV_2018_paper.pdf\n\n\n\n\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsseung0703%2Fsskd_svd","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fsseung0703%2Fsskd_svd","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsseung0703%2Fsskd_svd/lists"}