</> Code with Papers by Syntology

Datasets › WebVision

WebVision

Introduced by Wen Li et al. in WebVision Database: Visual Learning and Understanding from Web Data archive 2025-07-28

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.

The 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.

HomepageBenchmarksPapers JSON

Benchmarks archive 2025-07-28

All 4 leaderboards whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

First row (archive order)PaperCode
Image Classification mini WebVision 1.0 LRA-diffusion (CLIP ViT) Top-1 Accuracy 84.16 Label-Retrieval-Augmented Diffusion Models for Learning... puar-playground/lra-diffusion 47 Compare
Image Classification WebVision-1000 MAM (ViT-B/16) Top-1 Accuracy 83.6 Improving Image Recognition by Retrieving from Web-Scale... — 16 Compare
Image Classification WebVision PropMix (Ours) Top 1 Accuracy 78.84 PropMix: Hard Sample Filtering and Proportional MixUp... filipe-research/propmix 2 Compare
Learning with noisy labels mini WebVision 1.0 ILL Top 1 Accuracy 79.37 Imprecise Label Learning: A Unified Framework for... hhhhhhao/general-framework-weak-supervision 1 Compare

Papers archive 2025-07-28

30 shown of 51 papers with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 179. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

DateSamples run Syntology
PSSCL: A progressive sample selection framework with contrastive loss designed for noisy labels 1 2 18 Dec 2024 not harvested
Label-Retrieval-Augmented Diffusion Models for Learning from Noisy Labels 1 1 31 May 2023 ran 7 of 9 samples (2 unverified)
Imprecise Label Learning: A Unified Framework for Learning with Various Imprecise Label Configurations 1 1 22 May 2023 not harvested
Improving Image Recognition by Retrieving from Web-Scale Image-Text Data 0 1 11 Apr 2023 not harvested
Twin Contrastive Learning with Noisy Labels 1 1 13 Mar 2023 not harvested
Class Prototype-based Cleaner for Label Noise Learning 1 1 21 Dec 2022 not harvested
Dynamic Loss For Robust Learning 1 1 22 Nov 2022 not harvested
Bootstrapping the Relationship Between Images and Their Clean and Noisy Labels 1 1 17 Oct 2022 not harvested
Centrality and Consistency: Two-Stage Clean Samples Identification for Learning with Instance-Dependent Noisy Labels 1 1 29 Jul 2022 ran 4 of 9 samples (5 unverified)
Selective-Supervised Contrastive Learning with Noisy Labels 1 1 8 Mar 2022 ran 1 of 1 samples (0 unverified; 1 pointer-only for licence)
CMW-Net: Learning a Class-Aware Sample Weighting Mapping for Robust Deep Learning 1 3 11 Feb 2022 ran 2 of 3 samples (1 unverified; 3 pointer-only for licence)
Learning with Neighbor Consistency for Noisy Labels 1 5 4 Feb 2022 ran 4 of 4 samples (0 unverified)
Two Wrongs Don't Make a Right: Combating Confirmation Bias in Learning with Label Noise 0 1 6 Dec 2021 not harvested
Hard Sample Aware Noise Robust Learning for Histopathology Image Classification 1 1 5 Dec 2021 not harvested
Sample Prior Guided Robust Model Learning to Suppress Noisy Labels 1 1 2 Dec 2021 not harvested
Multi-label Iterated Learning for Image Classification with Label Ambiguity 0 2 23 Nov 2021 not harvested
CoDiM: Learning with Noisy Labels via Contrastive Semi-Supervised Learning 0 2 23 Nov 2021 not harvested
SSR: An Efficient and Robust Framework for Learning with Unknown Label Noise 1 1 22 Nov 2021 not harvested
PropMix: Hard Sample Filtering and Proportional MixUp for Learning with Noisy Labels 1 1 22 Oct 2021 not harvested
Robust Temporal Ensembling for Learning with Noisy Labels 0 1 29 Sep 2021 not harvested
Robust Long-Tailed Learning under Label Noise 0 1 26 Aug 2021 not harvested
NGC: A Unified Framework for Learning with Open-World Noisy Data 0 1 25 Aug 2021 not harvested
Confidence Adaptive Regularization for Deep Learning with Noisy Labels 0 1 18 Aug 2021 not harvested
Sample Selection with Uncertainty of Losses for Learning with Noisy Labels 0 1 1 Jun 2021 not harvested
Correlated Input-Dependent Label Noise in Large-Scale Image Classification 0 1 19 May 2021 not harvested
Generalized Jensen-Shannon Divergence Loss for Learning with Noisy Labels 1 1 10 May 2021 ran 4 of 5 samples (1 unverified; 5 pointer-only for licence)
Faster Meta Update Strategy for Noise-Robust Deep Learning 1 1 30 Apr 2021 ran 2 of 7 samples (5 unverified; 7 pointer-only for licence)
Contrast to Divide: Self-Supervised Pre-Training for Learning with Noisy Labels 1 1 25 Mar 2021 ran 6 of 7 samples (1 unverified)
ScanMix: Learning from Severe Label Noise via Semantic Clustering and Semi-Supervised Learning 1 1 21 Mar 2021 not harvested
LongReMix: Robust Learning with High Confidence Samples in a Noisy Label Environment 1 1 6 Mar 2021 not harvested

The full list of 51 is in the JSON twin.

Dataset loaders archive 2025-07-28

No loader listed in the archive.

Tasks archive 2025-07-28

Image ClassificationLearning with noisy labels

License archive 2025-07-28

Custom

Modalities archive 2025-07-28

Images

Languages archive 2025-07-28

No language tagged.

Variants archive 2025-07-28

  • WebVision-1000
  • WebVision
  • mini WebVision 1.0

3 variant names, as the archive lists them.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections