{"url":"/dataset/webvision-database","name":"WebVision","full_name":null,"description_markdown":"The WebVision dataset is designed to facilitate the research on learning visual representation from noisy web data. It is a large scale web images dataset that contains more than 2.4 million of images crawled from the Flickr website and Google Images search. \r\n\r\nThe same 1,000 concepts as the ILSVRC 2012 dataset are used for querying images, such that a bunch of existing approaches can be directly investigated and compared to the models trained from the ILSVRC 2012 dataset, and also makes it possible to study the dataset bias issue in the large scale scenario. The textual information accompanied with those images (e.g., caption, user tags, or description) are also provided as additional meta information. A validation set contains 50,000 images (50 images per category) is provided to facilitate the algorithmic development.","description_withheld":null,"homepage":"https://data.vision.ee.ethz.ch/cvl/webvision/dataset2017.html","introduced_date":null,"introduced_date_note":null,"introduced_by":{"paper":"/paper/webvision-database-visual-learning-and","title":"WebVision Database: Visual Learning and Understanding from Web Data","first_author":"Wen Li","url":null},"license":{"name":"Custom","url":"https://data.vision.ee.ethz.ch/cvl/webvision/download.html"},"modalities":[{"name":"Images","url":"/datasets/modality/images"}],"tasks":[{"name":"Image Classification","url":"/task/image-classification","datasets_with_task":"/datasets/task/image-classification"},{"name":"Learning with noisy labels","url":"/task/learning-with-noisy-labels","datasets_with_task":"/datasets/task/learning-with-noisy-labels"}],"languages":[],"variants":["WebVision-1000","WebVision","mini WebVision 1.0"],"data_loaders":[],"num_papers_in_archive":179,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/image-classification-on-mini-webvision-1-0","task":"Image Classification","dataset_variant":"mini WebVision 1.0","rows":47,"metrics":["Top-1 Accuracy","Top-5 Accuracy","ImageNet Top-1 Accuracy","ImageNet Top-5 Accuracy"],"first_row_in_archive_order":{"model":"LRA-diffusion (CLIP ViT)","paper":"/paper/label-retrieval-augmented-diffusion-models-1","metrics":{"ImageNet Top-1 Accuracy":"82.56","Top-1 Accuracy":"84.16"},"code_links":[{"title":"puar-playground/lra-diffusion","url":"https://github.com/puar-playground/lra-diffusion"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/image-classification-on-webvision-1000","task":"Image Classification","dataset_variant":"WebVision-1000","rows":16,"metrics":["Top-1 Accuracy","Top-5 Accuracy","ImageNet Top-1 Accuracy","ImageNet Top-5 Accuracy"],"first_row_in_archive_order":{"model":"MAM (ViT-B/16)","paper":"/paper/improving-image-recognition-by-retrieving","metrics":{"Top-1 Accuracy":"83.6"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/image-classification-on-webvision","task":"Image Classification","dataset_variant":"WebVision","rows":2,"metrics":["Top 1 Accuracy","Top 5 Accuracy"],"first_row_in_archive_order":{"model":"PropMix (Ours)","paper":"/paper/propmix-hard-sample-filtering-and","metrics":{"Top 1 Accuracy":"78.84"},"code_links":[{"title":"filipe-research/propmix","url":"https://github.com/filipe-research/propmix"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/learning-with-noisy-labels-on-mini-webvision","task":"Learning with noisy labels","dataset_variant":"mini WebVision 1.0","rows":1,"metrics":["Top 1 Accuracy"],"first_row_in_archive_order":{"model":"ILL","paper":"/paper/imprecise-label-learning-a-unified-framework","metrics":{"Top 1 Accuracy":"79.37"},"code_links":[{"title":"hhhhhhao/general-framework-weak-supervision","url":"https://github.com/hhhhhhao/general-framework-weak-supervision"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/psscl-a-progressive-sample-selection","title":"PSSCL: A progressive sample selection framework with contrastive loss designed for noisy labels","date":"2024-12-18","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/label-retrieval-augmented-diffusion-models-1","title":"Label-Retrieval-Augmented Diffusion Models for Learning from Noisy Labels","date":"2023-05-31","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":9,"samples_ran":7,"samples_unverified":2,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/imprecise-label-learning-a-unified-framework","title":"Imprecise Label Learning: A Unified Framework for Learning with Various Imprecise Label Configurations","date":"2023-05-22","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/improving-image-recognition-by-retrieving","title":"Improving Image Recognition by Retrieving from Web-Scale Image-Text Data","date":"2023-04-11","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/twin-contrastive-learning-with-noisy-labels","title":"Twin Contrastive Learning with Noisy Labels","date":"2023-03-13","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/class-prototype-based-cleaner-for-label-noise","title":"Class Prototype-based Cleaner for Label Noise Learning","date":"2022-12-21","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/dynamic-loss-for-robust-learning","title":"Dynamic Loss For Robust Learning","date":"2022-11-22","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/bootstrapping-the-relationship-between-images","title":"Bootstrapping the Relationship Between Images and Their Clean and Noisy Labels","date":"2022-10-17","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/centrality-and-consistency-two-stage-clean","title":"Centrality and Consistency: Two-Stage Clean Samples Identification for Learning with Instance-Dependent Noisy Labels","date":"2022-07-29","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":9,"samples_ran":4,"samples_unverified":5,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/selective-supervised-contrastive-learning","title":"Selective-Supervised Contrastive Learning with Noisy Labels","date":"2022-03-08","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":1,"samples_ran":1,"samples_unverified":0,"pointer_only_for_licence":1,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/cmw-net-learning-a-class-aware-sample","title":"CMW-Net: Learning a Class-Aware Sample Weighting Mapping for Robust Deep Learning","date":"2022-02-11","rows_on_this_dataset":3,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":3,"samples_ran":2,"samples_unverified":1,"pointer_only_for_licence":3,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/learning-with-neighbor-consistency-for-noisy-1","title":"Learning with Neighbor Consistency for Noisy Labels","date":"2022-02-04","rows_on_this_dataset":5,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":4,"samples_ran":4,"samples_unverified":0,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/two-wrongs-don-t-make-a-right-combating","title":"Two Wrongs Don't Make a Right: Combating Confirmation Bias in Learning with Label Noise","date":"2021-12-06","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/hard-sample-aware-noise-robust-learning-for","title":"Hard Sample Aware Noise Robust Learning for Histopathology Image Classification","date":"2021-12-05","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/sample-prior-guided-robust-model-learning-to","title":"Sample Prior Guided Robust Model Learning to Suppress Noisy Labels","date":"2021-12-02","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/multi-label-iterated-learning-for-image","title":"Multi-label Iterated Learning for Image Classification with Label Ambiguity","date":"2021-11-23","rows_on_this_dataset":2,"code_links":0,"syntology":null},{"paper":"/paper/codim-learning-with-noisy-labels-via","title":"CoDiM: Learning with Noisy Labels via Contrastive Semi-Supervised Learning","date":"2021-11-23","rows_on_this_dataset":2,"code_links":0,"syntology":null},{"paper":"/paper/s3-supervised-self-supervised-learning-under-1","title":"SSR: An Efficient and Robust Framework for Learning with Unknown Label Noise","date":"2021-11-22","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/propmix-hard-sample-filtering-and","title":"PropMix: Hard Sample Filtering and Proportional MixUp for Learning with Noisy Labels","date":"2021-10-22","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/robust-temporal-ensembling-for-learning-with","title":"Robust Temporal Ensembling for Learning with Noisy Labels","date":"2021-09-29","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/robust-long-tailed-learning-under-label-noise","title":"Robust Long-Tailed Learning under Label Noise","date":"2021-08-26","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/ngc-a-unified-framework-for-learning-with","title":"NGC: A Unified Framework for Learning with Open-World Noisy Data","date":"2021-08-25","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/confidence-adaptive-regularization-for-deep","title":"Confidence Adaptive Regularization for Deep Learning with Noisy Labels","date":"2021-08-18","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/sample-selection-with-uncertainty-of-losses","title":"Sample Selection with Uncertainty of Losses for Learning with Noisy Labels","date":"2021-06-01","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/correlated-input-dependent-label-noise-in","title":"Correlated Input-Dependent Label Noise in Large-Scale Image Classification","date":"2021-05-19","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/generalized-jensen-shannon-divergence-loss","title":"Generalized Jensen-Shannon Divergence Loss for Learning with Noisy Labels","date":"2021-05-10","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":5,"samples_ran":4,"samples_unverified":1,"pointer_only_for_licence":5,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/faster-meta-update-strategy-for-noise-robust","title":"Faster Meta Update Strategy for Noise-Robust Deep Learning","date":"2021-04-30","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":7,"samples_ran":2,"samples_unverified":5,"pointer_only_for_licence":7,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/contrast-to-divide-self-supervised-pre-1","title":"Contrast to Divide: Self-Supervised Pre-Training for Learning with Noisy Labels","date":"2021-03-25","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":7,"samples_ran":6,"samples_unverified":1,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/scanmix-learning-from-severe-label-noise-via","title":"ScanMix: Learning from Severe Label Noise via Semantic Clustering and Semi-Supervised Learning","date":"2021-03-21","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/longremix-robust-learning-with-high","title":"LongReMix: Robust Learning with High Confidence Samples in a Noisy Label Environment","date":"2021-03-06","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/winning-ticket-in-noisy-image-classification","title":"FINE Samples for Learning with Noisy Labels","date":"2021-02-23","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":7,"samples_ran":5,"samples_unverified":2,"pointer_only_for_licence":7,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/robust-early-learning-hindering-the","title":"Robust early-learning: Hindering the memorization of noisy labels","date":"2021-01-01","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/multi-objective-interpolation-training-for","title":"Multi-Objective Interpolation Training for Robustness to Label Noise","date":"2020-12-08","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/coresets-for-robust-training-of-neural","title":"Coresets for Robust Training of Neural Networks against Noisy Labels","date":"2020-11-15","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/webly-supervised-image-classification-with-1","title":"Webly Supervised Image Classification with Metadata: Automatic Noisy Label Correction via Visual-Semantic Graph","date":"2020-10-12","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/noisy-concurrent-training-for-efficient","title":"Noisy Concurrent Training for Efficient Learning under Label Noise","date":"2020-09-17","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":6,"samples_ran":5,"samples_unverified":1,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/mopro-webly-supervised-learning-with-momentum","title":"MoPro: Webly Supervised Learning with Momentum Prototypes","date":"2020-09-17","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":1,"samples_ran":0,"samples_unverified":1,"pointer_only_for_licence":1,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/webly-supervised-image-classification-with","title":"Webly Supervised Image Classification with Self-Contained Confidence","date":"2020-08-27","rows_on_this_dataset":1,"code_links":4,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":4,"samples_ran":0,"samples_unverified":4,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/early-learning-regularization-prevents","title":"Early-Learning Regularization Prevents Memorization of Noisy Labels","date":"2020-06-30","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":1,"samples_ran":1,"samples_unverified":0,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/heteroskedastic-and-imbalanced-deep-learning","title":"Heteroskedastic and Imbalanced Deep Learning with Adaptive Regularization","date":"2020-06-29","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":2,"samples_ran":0,"samples_unverified":2,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/normalized-loss-functions-for-deep-learning","title":"Normalized Loss Functions for Deep Learning with Noisy Labels","date":"2020-06-24","rows_on_this_dataset":2,"code_links":4,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":12,"samples_ran":1,"samples_unverified":11,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/robust-and-on-the-fly-dataset-denoising-for","title":"Robust and On-the-fly Dataset Denoising for Image Classification","date":"2020-03-24","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/dividemix-learning-with-noisy-labels-as-semi-1","title":"DivideMix: Learning with Noisy Labels as Semi-supervised Learning","date":"2020-02-18","rows_on_this_dataset":3,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":2,"samples_ran":0,"samples_unverified":2,"pointer_only_for_licence":2,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/synthetic-vs-real-deep-learning-on-controlled-1","title":"Beyond Synthetic Noise: Deep Learning on Controlled Noisy Labels","date":"2019-11-21","rows_on_this_dataset":2,"code_links":3,"syntology":null},{"paper":"/paper/protonet-learning-from-web-data-with-memory","title":"Learning from Web Data with Self-Organizing Memory Module","date":"2019-06-28","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/understanding-and-utilizing-deep-neural","title":"Understanding and Utilizing Deep Neural Networks Trained with Noisy Labels","date":"2019-05-13","rows_on_this_dataset":1,"code_links":3,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":3,"samples_ran":3,"samples_unverified":0,"pointer_only_for_licence":3,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/curriculumnet-weakly-supervised-learning-from","title":"CurriculumNet: Weakly Supervised Learning from Large-Scale Web Images","date":"2018-08-03","rows_on_this_dataset":2,"code_links":2,"syntology":null},{"paper":"/paper/dimensionality-driven-learning-with-noisy","title":"Dimensionality-Driven Learning with Noisy Labels","date":"2018-06-07","rows_on_this_dataset":1,"code_links":3,"syntology":null},{"paper":"/paper/co-teaching-robust-training-of-deep-neural","title":"Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels","date":"2018-04-18","rows_on_this_dataset":1,"code_links":5,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":7,"samples_ran":7,"samples_unverified":0,"pointer_only_for_licence":7,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/mentornet-learning-data-driven-curriculum-for","title":"MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels","date":"2017-12-14","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/making-deep-neural-networks-robust-to-label","title":"Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approach","date":"2016-09-13","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":2,"samples_ran":2,"samples_unverified":0,"pointer_only_for_licence":2,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":19,"samples_harvested":92,"samples_ran":54,"samples_unverified":38,"pointer_only_for_licence":38,"papers_with_no_sample_that_ran":4,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}