{"url":"/sota/image-classification-on-omnibenchmark","task":{"name":"Image Classification","url":"/task/image-classification","note":null},"dataset":{"name":"OmniBenchmark","url":"/dataset/omnibenchmark"},"category":"Computer Vision","categories":["Adversarial","Computer Vision"],"category_note":null,"description":"**Image Classification** is a fundamental task in vision recognition that aims to understand and categorize an image as a whole under a specific label. Unlike [object detection](/task/object-detection), which involves classification and location of multiple objects within an image, image classification typically pertains to single-object images. When the classification becomes highly detailed or reaches instance-level, it is often referred to as [image retrieval](/task/image-retrieval), which also involves finding similar images in a large database.\r\n\r\n\r\n<span class=\"description-source\">Source: [Metamorphic Testing for Object Detection Systems ](https://arxiv.org/abs/1912.12162)</span>","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["Average Top-1 Accuracy"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Average Top-1 Accuracy":"higher"}},"counts":{"rows":22,"rows_with_code":22,"rows_with_paper_page":22,"rows_dated":22,"rows_using_additional_data":7},"rows":[{"rank_in_archive_order":1,"model":"NOAH-ViTB/16","metrics":{"Average Top-1 Accuracy":"47.6"},"uses_additional_data":true,"paper_date":"2022-06-09","paper":"/paper/neural-prompt-search","paper_url":"https://arxiv.org/abs/2206.04673v2","paper_title":"Neural Prompt Search","code":"https://github.com/ZhangYuanhan-AI/NOAH","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":3,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":2,"model":"SwinTransformer","metrics":{"Average Top-1 Accuracy":"46.4"},"uses_additional_data":false,"paper_date":"2021-03-25","paper":"/paper/swin-transformer-hierarchical-vision","paper_url":"https://arxiv.org/abs/2103.14030v2","paper_title":"Swin Transformer: Hierarchical Vision Transformer using Shifted Windows","code":"https://github.com/huggingface/transformers","n_code_links":80,"syntology":{"n_ran":108,"n_unverified":99,"n_samples":207,"n_pointer_only_licence":43}},{"rank_in_archive_order":3,"model":"Bamboo-R50","metrics":{"Average Top-1 Accuracy":"45.4"},"uses_additional_data":true,"paper_date":"2022-03-15","paper":"/paper/bamboo-building-mega-scale-vision-dataset","paper_url":"https://arxiv.org/abs/2203.07845v2","paper_title":"Bamboo: Building Mega-Scale Vision Dataset Continually with Human-Machine Synergy","code":"https://github.com/zhangyuanhan-ai/bamboo","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":4,"model":"Adapter-ViTB/16","metrics":{"Average Top-1 Accuracy":"44.5"},"uses_additional_data":true,"paper_date":"2019-02-02","paper":"/paper/parameter-efficient-transfer-learning-for-nlp","paper_url":"https://arxiv.org/abs/1902.00751v2","paper_title":"Parameter-Efficient Transfer Learning for NLP","code":"https://github.com/Adapter-Hub/adapter-transformers","n_code_links":17,"syntology":{"n_ran":8,"n_unverified":14,"n_samples":22,"n_pointer_only_licence":3}},{"rank_in_archive_order":5,"model":"CLIP-RN50","metrics":{"Average Top-1 Accuracy":"42.1"},"uses_additional_data":true,"paper_date":"2021-02-26","paper":"/paper/learning-transferable-visual-models-from","paper_url":"https://arxiv.org/abs/2103.00020v1","paper_title":"Learning Transferable Visual Models From Natural Language Supervision","code":"https://github.com/openai/CLIP","n_code_links":82,"syntology":{"n_ran":16,"n_unverified":4,"n_samples":20,"n_pointer_only_licence":16}},{"rank_in_archive_order":6,"model":"IG-1B","metrics":{"Average Top-1 Accuracy":"40.4"},"uses_additional_data":true,"paper_date":"2019-05-02","paper":"/paper/billion-scale-semi-supervised-learning-for","paper_url":"http://arxiv.org/abs/1905.00546v1","paper_title":"Billion-scale semi-supervised learning for image classification","code":"https://github.com/facebookresearch/semi-supervised-ImageNet1K-models","n_code_links":4,"syntology":null},{"rank_in_archive_order":7,"model":"BiT-M","metrics":{"Average Top-1 Accuracy":"40.4"},"uses_additional_data":true,"paper_date":"2019-12-24","paper":"/paper/large-scale-learning-of-general-visual","paper_url":"https://arxiv.org/abs/1912.11370v3","paper_title":"Big Transfer (BiT): General Visual Representation Learning","code":"https://github.com/google-research/big_transfer","n_code_links":9,"syntology":{"n_ran":3,"n_unverified":7,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":8,"model":"DINO","metrics":{"Average Top-1 Accuracy":"38.9"},"uses_additional_data":false,"paper_date":"2021-04-29","paper":"/paper/emerging-properties-in-self-supervised-vision","paper_url":"https://arxiv.org/abs/2104.14294v2","paper_title":"Emerging Properties in Self-Supervised Vision Transformers","code":"https://github.com/facebookresearch/dino","n_code_links":32,"syntology":{"n_ran":5,"n_unverified":15,"n_samples":20,"n_pointer_only_licence":2}},{"rank_in_archive_order":9,"model":"SwAV","metrics":{"Average Top-1 Accuracy":"38.3"},"uses_additional_data":false,"paper_date":"2020-06-17","paper":"/paper/unsupervised-learning-of-visual-features-by","paper_url":"https://arxiv.org/abs/2006.09882v5","paper_title":"Unsupervised Learning of Visual Features by Contrasting Cluster Assignments","code":"https://github.com/open-mmlab/mmdetection","n_code_links":18,"syntology":{"n_ran":13,"n_unverified":4,"n_samples":17,"n_pointer_only_licence":6}},{"rank_in_archive_order":10,"model":"ResNet-101","metrics":{"Average Top-1 Accuracy":"37.4"},"uses_additional_data":false,"paper_date":"2015-12-10","paper":"/paper/deep-residual-learning-for-image-recognition","paper_url":"http://arxiv.org/abs/1512.03385v1","paper_title":"Deep Residual Learning for Image Recognition","code":"https://github.com/tensorflow/models/tree/master/research/deeplab","n_code_links":484,"syntology":{"n_ran":230,"n_unverified":147,"n_samples":377,"n_pointer_only_licence":187}},{"rank_in_archive_order":11,"model":"MEAL-V2","metrics":{"Average Top-1 Accuracy":"36.6"},"uses_additional_data":false,"paper_date":"2020-09-17","paper":"/paper/meal-v2-boosting-vanilla-resnet-50-to-80-top","paper_url":"https://arxiv.org/abs/2009.08453v2","paper_title":"MEAL V2: Boosting Vanilla ResNet-50 to 80%+ Top-1 Accuracy on ImageNet without Tricks","code":"https://github.com/szq0214/MEAL-V2","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"MoPro-V2","metrics":{"Average Top-1 Accuracy":"36.1"},"uses_additional_data":true,"paper_date":"2020-09-17","paper":"/paper/mopro-webly-supervised-learning-with-momentum","paper_url":"https://arxiv.org/abs/2009.07995v1","paper_title":"MoPro: Webly Supervised Learning with Momentum Prototypes","code":"https://github.com/salesforce/MoPro","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":13,"model":"EfficientNetB4","metrics":{"Average Top-1 Accuracy":"35.8"},"uses_additional_data":false,"paper_date":"2019-05-28","paper":"/paper/efficientnet-rethinking-model-scaling-for","paper_url":"https://arxiv.org/abs/1905.11946v5","paper_title":"EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks","code":"https://github.com/ultralytics/yolov5","n_code_links":144,"syntology":{"n_ran":171,"n_unverified":131,"n_samples":302,"n_pointer_only_licence":112}},{"rank_in_archive_order":14,"model":"MoCoV2","metrics":{"Average Top-1 Accuracy":"34.8"},"uses_additional_data":false,"paper_date":"2019-11-13","paper":"/paper/momentum-contrast-for-unsupervised-visual","paper_url":"https://arxiv.org/abs/1911.05722v3","paper_title":"Momentum Contrast for Unsupervised Visual Representation Learning","code":"https://github.com/open-mmlab/mmdetection","n_code_links":44,"syntology":{"n_ran":26,"n_unverified":16,"n_samples":42,"n_pointer_only_licence":16}},{"rank_in_archive_order":15,"model":"ResNet-50","metrics":{"Average Top-1 Accuracy":"34.3"},"uses_additional_data":false,"paper_date":"2015-12-10","paper":"/paper/deep-residual-learning-for-image-recognition","paper_url":"http://arxiv.org/abs/1512.03385v1","paper_title":"Deep Residual Learning for Image Recognition","code":"https://github.com/tensorflow/models/tree/master/research/deeplab","n_code_links":484,"syntology":{"n_ran":230,"n_unverified":147,"n_samples":377,"n_pointer_only_licence":187}},{"rank_in_archive_order":16,"model":"InceptionV4","metrics":{"Average Top-1 Accuracy":"32.3"},"uses_additional_data":false,"paper_date":"2016-02-23","paper":"/paper/inception-v4-inception-resnet-and-the-impact","paper_url":"http://arxiv.org/abs/1602.07261v2","paper_title":"Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning","code":"https://github.com/tensorflow/models","n_code_links":87,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":17,"model":"MLP-Mixer","metrics":{"Average Top-1 Accuracy":"32.2"},"uses_additional_data":false,"paper_date":"2021-05-04","paper":"/paper/mlp-mixer-an-all-mlp-architecture-for-vision","paper_url":"https://arxiv.org/abs/2105.01601v4","paper_title":"MLP-Mixer: An all-MLP Architecture for Vision","code":"https://github.com/labmlai/annotated_deep_learning_paper_implementations","n_code_links":49,"syntology":{"n_ran":106,"n_unverified":28,"n_samples":134,"n_pointer_only_licence":36}},{"rank_in_archive_order":18,"model":"Manifold","metrics":{"Average Top-1 Accuracy":"31.6"},"uses_additional_data":false,"paper_date":"2018-06-13","paper":"/paper/manifold-mixup-better-representations-by","paper_url":"https://arxiv.org/abs/1806.05236v7","paper_title":"Manifold Mixup: Better Representations by Interpolating Hidden States","code":"https://github.com/Westlake-AI/openmixup","n_code_links":12,"syntology":{"n_ran":2,"n_unverified":9,"n_samples":11,"n_pointer_only_licence":5}},{"rank_in_archive_order":19,"model":"CutMix","metrics":{"Average Top-1 Accuracy":"31.1"},"uses_additional_data":false,"paper_date":"2019-05-13","paper":"/paper/cutmix-regularization-strategy-to-train","paper_url":"https://arxiv.org/abs/1905.04899v2","paper_title":"CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":30,"syntology":{"n_ran":17,"n_unverified":7,"n_samples":24,"n_pointer_only_licence":5}},{"rank_in_archive_order":20,"model":"ReLabel","metrics":{"Average Top-1 Accuracy":"30.8"},"uses_additional_data":false,"paper_date":"2021-01-13","paper":"/paper/re-labeling-imagenet-from-single-to-multi","paper_url":"https://arxiv.org/abs/2101.05022v2","paper_title":"Re-labeling ImageNet: from Single to Multi-Labels, from Global to Localized Labels","code":"https://github.com/naver-ai/relabel_imagenet","n_code_links":2,"syntology":null},{"rank_in_archive_order":21,"model":"MAE","metrics":{"Average Top-1 Accuracy":"30.6"},"uses_additional_data":false,"paper_date":"2021-11-11","paper":"/paper/masked-autoencoders-are-scalable-vision","paper_url":"https://arxiv.org/abs/2111.06377v2","paper_title":"Masked Autoencoders Are Scalable Vision Learners","code":"https://github.com/facebookresearch/mae","n_code_links":58,"syntology":{"n_ran":71,"n_unverified":66,"n_samples":137,"n_pointer_only_licence":73}},{"rank_in_archive_order":22,"model":"BeiT","metrics":{"Average Top-1 Accuracy":"30.1"},"uses_additional_data":false,"paper_date":"2021-06-15","paper":"/paper/beit-bert-pre-training-of-image-transformers","paper_url":"https://arxiv.org/abs/2106.08254v2","paper_title":"BEiT: BERT Pre-Training of Image Transformers","code":"https://github.com/huggingface/transformers","n_code_links":14,"syntology":{"n_ran":6,"n_unverified":5,"n_samples":11,"n_pointer_only_licence":0}}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,264 of the 9,581 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9581,"papers_checked":6264,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":3316},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":19,"rows_with_any_sample_ran":18,"distinct_papers_with_graph_line":18,"distinct_papers_with_any_sample_ran":17,"samples_over_distinct_papers":{"n_ran":792,"n_unverified":556,"n_samples":1348,"n_pointer_only_licence":508,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":1022,"n_unverified":703,"n_samples":1725,"n_pointer_only_licence":695,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}