{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/robustness-of-conditional-gans-to-noisy","title":"Robustness of Conditional GANs to Noisy Labels","arxiv_id":"1811.03205","date":"2018-11-08","proceeding":"NeurIPS 2018 12","authors":["Kiran Koshy Thekumparampil","Ashish Khetan","Zinan Lin","Sewoong Oh"],"abstract":"We study the problem of learning conditional generators from noisy labeled\nsamples, where the labels are corrupted by random noise. A standard training of\nconditional GANs will not only produce samples with wrong labels, but also\ngenerate poor quality samples. We consider two scenarios, depending on whether\nthe noise model is known or not. When the distribution of the noise is known,\nwe introduce a novel architecture which we call Robust Conditional GAN (RCGAN).\nThe main idea is to corrupt the label of the generated sample before feeding to\nthe adversarial discriminator, forcing the generator to produce samples with\nclean labels. This approach of passing through a matching noisy channel is\njustified by corresponding multiplicative approximation bounds between the loss\nof the RCGAN and the distance between the clean real distribution and the\ngenerator distribution. This shows that the proposed approach is robust, when\nused with a carefully chosen discriminator architecture, known as projection\ndiscriminator. When the distribution of the noise is not known, we provide an\nextension of our architecture, which we call RCGAN-U, that learns the noise\nmodel simultaneously while training the generator. We show experimentally on\nMNIST and CIFAR-10 datasets that both the approaches consistently improve upon\nbaseline approaches, and RCGAN-U closely matches the performance of RCGAN.","url_abs":"http://arxiv.org/abs/1811.03205v1","url_pdf":"http://arxiv.org/pdf/1811.03205v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"robustness-of-conditional-gans-to-noisy","repo_url":"https://github.com/POLane16/Robust-Conditional-GAN","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"robustness-of-conditional-gans-to-noisy","repo_url":"https://github.com/takuhirok/rGAN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1811.03205","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1811.03205"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/takuhirok/rGAN","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/POLane16/Robust-Conditional-GAN","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_honours":1,"unverified":12},"by_repo_kind":{"official":{"samples":13,"ran":1,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"40125e1efb67fa92","entry":"conv_out_size_same","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"mnist/model.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/mnist/model.py","link_basis":"harvester_set","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"40125e1efb67fa92"}},{"code_sha256_prefix":"2ec0891e6548e951","entry":"MultiScaleSSIM","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"cifar10/common/msssim.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/cifar10/common/msssim.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"2ec0891e6548e951"}},{"code_sha256_prefix":"97a17fbee43a92bd","entry":"conv_cond_concat","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"mnist/ops.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/mnist/ops.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"97a17fbee43a92bd"}},{"code_sha256_prefix":"fc22db7fa7e0e135","entry":"custom_crop","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"cifar10/common/misc.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/cifar10/common/misc.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"fc22db7fa7e0e135"}},{"code_sha256_prefix":"f43eeb47efdaf9c4","entry":"deconv2d","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"mnist/ops.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/mnist/ops.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f43eeb47efdaf9c4"}},{"code_sha256_prefix":"a0b95d6b4e00d081","entry":"get_image","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"mnist/utils.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/mnist/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a0b95d6b4e00d081"}},{"code_sha256_prefix":"c69f8aa044146a22","entry":"get_image","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"cifar10/common/misc.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/cifar10/common/misc.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c69f8aa044146a22"}},{"code_sha256_prefix":"f2972a3201c29e83","entry":"imread","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"mnist/utils.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/mnist/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f2972a3201c29e83"}},{"code_sha256_prefix":"e98109fb35cf54c3","entry":"nonlinearity","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"cifar10/gan_resnet.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/cifar10/gan_resnet.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e98109fb35cf54c3"}},{"code_sha256_prefix":"7fdc6f75b649fe37","entry":"preds2score","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"cifar10/common/inception/inception_score.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/cifar10/common/inception/inception_score.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"7fdc6f75b649fe37"}},{"code_sha256_prefix":"3218db3da6c1ad6c","entry":"save_images","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"mnist/utils.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/mnist/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"3218db3da6c1ad6c"}},{"code_sha256_prefix":"805d7edd87867aad","entry":"spectral_normed_weight","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"mnist/sn.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/mnist/sn.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"805d7edd87867aad"}},{"code_sha256_prefix":"b67362dfc6797565","entry":"transform","repo":"POLane16/Robust-Conditional-GAN","repo_kind":"official","path":"cifar10/common/misc.py","file_url":"https://github.com/POLane16/Robust-Conditional-GAN/blob/HEAD/cifar10/common/misc.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b67362dfc6797565"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}