{"id":19839851,"url":"https://github.com/qdata/deepcloak","last_synced_at":"2025-05-01T19:30:25.966Z","repository":{"id":83652133,"uuid":"92061419","full_name":"QData/DeepCloak","owner":"QData","description":"ICLR16: DeepCloak: Masking Deep Neural Network Models for Robustness Against Adversarial Samples","archived":false,"fork":false,"pushed_at":"2019-08-28T16:29:08.000Z","size":11,"stargazers_count":6,"open_issues_count":0,"forks_count":1,"subscribers_count":4,"default_branch":"master","last_synced_at":"2025-04-06T17:05:57.561Z","etag":null,"topics":["adversarial-samples","mask","residual-networks","robustness"],"latest_commit_sha":null,"homepage":"https://arxiv.org/abs/1702.06763","language":"Lua","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/QData.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,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2017-05-22T14:13:51.000Z","updated_at":"2021-03-04T09:19:56.000Z","dependencies_parsed_at":"2023-03-12T18:59:24.950Z","dependency_job_id":null,"html_url":"https://github.com/QData/DeepCloak","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/QData%2FDeepCloak","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/QData%2FDeepCloak/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/QData%2FDeepCloak/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/QData%2FDeepCloak/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/QData","download_url":"https://codeload.github.com/QData/DeepCloak/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":251932525,"owners_count":21667159,"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":["adversarial-samples","mask","residual-networks","robustness"],"created_at":"2024-11-12T12:24:34.935Z","updated_at":"2025-05-01T19:30:25.960Z","avatar_url":"https://github.com/QData.png","language":"Lua","funding_links":[],"categories":[],"sub_categories":[],"readme":"# DeepCloak: Masking Deep Neural Network Models for Robustness Against Adversarial Samples\n\n## Environment: Torch7 + CUDNN\n\n\n\n## Reference \n\nat Workshop of ICLR16: \n\n### Title: DeepCloak: Masking Deep Neural Network Models for Robustness Against Adversarial Samples \n\n### [URL: https://arxiv.org/abs/1702.06763](https://arxiv.org/abs/1702.06763)\n\n```\n@article{GaoWQ17,\n  author    = {Ji Gao and\n               Beilun Wang and\n               Yanjun Qi},\n  title     = {DeepCloak: Masking {DNN} Models for robustness against adversarial\n               samples},\n  journal   = {CoRR},\n  volume    = {abs/1702.06763},\n  year      = {2017},\n  url       = {http://arxiv.org/abs/1702.06763},\n  archivePrefix = {arXiv},\n  eprint    = {1702.06763},\n  biburl    = {https://dblp.org/rec/bib/journals/corr/GaoWQ17},\n}\n```\n\n## Example:\n\nth removenode.lua -dataset resources/cifar10.t7 -model resources/model_res-164.t7 -layernum 8\n\n## Usage: \n\nth removenode.lua -model MODELADD -dataset DATASETADD -layernum LAYERNUM -std STD [-power POWER] [-gpu GPUNUM] \n\n * [MODELADD]: address of the model file \\n\n\n * [LAYERNUM]: number of the layer where the mask will be inserted after it\n\n * [POWER]: attack strength, epsilon in Fast Gradient Sign Method, default 10 \n\n * [GPUNUM]: number of GPU selected\n\n * [DATASETADD]: address of the dataset file\n\n * [STD]: the standard deviation of the dataset used in the preprocessing, required in the Adversarial Sample Generation\n\n## Dataset and models: Orginially from https://github.com/szagoruyko/wide-residual-networks\n\n* Download them from http://www.cs.virginia.edu/~jg6yd/resources/\n\n* Cifar-10: Whitened data of CIFAR-10, std = 17.067, which is selected as the default dataset\n\n* model_vgg_orig.t7: A pretrained model of VGG-16\n\n* model_res-164.t7: A pretrained model of the 152 layer residual network\n\n* Wide.t7: A pretrained model of wide residual network from https://github.com/szagoruyko/wide-residual-networks\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fqdata%2Fdeepcloak","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fqdata%2Fdeepcloak","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fqdata%2Fdeepcloak/lists"}