{"url":"/sota/metric-learning-on-cars196","task":{"name":"Metric Learning","url":"/task/metric-learning","note":null},"dataset":{"name":"CARS196","url":"/dataset/cars196"},"category":"Computer Vision","categories":["Computer Vision","Methodology"],"category_note":null,"description":"The goal of **Metric Learning** is to learn a representation function that maps objects into an embedded space. The distance in the embedded space should preserve the objects’ similarity — similar objects get close and dissimilar objects get far away. Various loss functions have been developed for Metric Learning. For example, the **contrastive loss** guides the objects from the same class to be mapped to the same point and those from different classes to be mapped to different points whose distances are larger than a margin. **Triplet loss** is also popular, which requires the distance between the anchor sample and the positive sample to be smaller than the distance between the anchor sample and the negative sample.\r\n\r\n\r\n<span class=\"description-source\">Source: [Road Network Metric Learning for Estimated Time of Arrival ](https://arxiv.org/abs/2006.13477)</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":["R@1"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"R@1":null}},"counts":{"rows":36,"rows_with_code":32,"rows_with_paper_page":36,"rows_dated":36,"rows_using_additional_data":20},"rows":[{"rank_in_archive_order":1,"model":"Unicom+ViT-L@336px","metrics":{"R@1":"98.2"},"uses_additional_data":true,"paper_date":"2023-04-12","paper":"/paper/unicom-universal-and-compact-representation","paper_url":"https://arxiv.org/abs/2304.05884v1","paper_title":"Unicom: Universal and Compact Representation Learning for Image Retrieval","code":"https://github.com/OML-Team/open-metric-learning","n_code_links":3,"syntology":{"n_ran":3,"n_unverified":3,"n_samples":6,"n_pointer_only_licence":6}},{"rank_in_archive_order":2,"model":"Hyp-DINO 8x8","metrics":{"R@1":"92.8"},"uses_additional_data":false,"paper_date":"2022-03-21","paper":"/paper/hyperbolic-vision-transformers-combining","paper_url":"https://arxiv.org/abs/2203.10833v2","paper_title":"Hyperbolic Vision Transformers: Combining Improvements in Metric Learning","code":"https://github.com/OML-Team/open-metric-learning","n_code_links":2,"syntology":{"n_ran":4,"n_unverified":2,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":3,"model":"NED","metrics":{"R@1":"91.5"},"uses_additional_data":false,"paper_date":"2020-06-08","paper":"/paper/calibrated-neighborhood-aware-confidence","paper_url":"https://arxiv.org/abs/2006.04935v1","paper_title":"Calibrated neighborhood aware confidence measure for deep metric learning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":4,"model":"ResNet-50 + Intra-Batch (ensemble of 5)","metrics":{"R@1":"91.5"},"uses_additional_data":false,"paper_date":"2021-02-15","paper":"/paper/learning-intra-batch-connections-for-deep","paper_url":"https://arxiv.org/abs/2102.07753v3","paper_title":"Learning Intra-Batch Connections for Deep Metric Learning","code":"https://github.com/dvl-tum/intra_batch","n_code_links":2,"syntology":null},{"rank_in_archive_order":5,"model":"ResNet-50 + AVSL","metrics":{"R@1":"91.5"},"uses_additional_data":true,"paper_date":"2022-03-28","paper":"/paper/attributable-visual-similarity-learning","paper_url":"https://arxiv.org/abs/2203.14932v1","paper_title":"Attributable Visual Similarity Learning","code":"https://github.com/zbr17/avsl","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":1,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":6,"model":"EfficientDML-VPTSP-G/512","metrics":{"R@1":"91.2"},"uses_additional_data":true,"paper_date":"2024-02-04","paper":"/paper/learning-semantic-proxies-from-visual-prompts","paper_url":"https://arxiv.org/abs/2402.02340v2","paper_title":"Learning Semantic Proxies from Visual Prompts for Parameter-Efficient Fine-Tuning in Deep Metric Learning","code":"https://github.com/noahsark/parameterefficient-dml","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":2,"n_samples":5,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"CCL (ResNet-50)","metrics":{"R@1":"91.02"},"uses_additional_data":false,"paper_date":"2023-08-01","paper":"/paper/center-contrastive-loss-for-metric-learning","paper_url":"https://arxiv.org/abs/2308.00458v1","paper_title":"Center Contrastive Loss for Metric Learning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":8,"model":"ResNet50 + Language","metrics":{"R@1":"90.2"},"uses_additional_data":true,"paper_date":"2022-03-16","paper":"/paper/integrating-language-guidance-into-vision","paper_url":"https://arxiv.org/abs/2203.08543v1","paper_title":"Integrating Language Guidance into Vision-based Deep Metric Learning","code":"https://github.com/explainableml/languageguidance_for_dml","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":1,"n_samples":5,"n_pointer_only_licence":0}},{"rank_in_archive_order":9,"model":"ResNet-50 + Metrix","metrics":{"R@1":"89.6"},"uses_additional_data":true,"paper_date":"2021-06-09","paper":"/paper/it-takes-two-to-tango-mixup-for-deep-metric","paper_url":"https://arxiv.org/abs/2106.04990v2","paper_title":"It Takes Two to Tango: Mixup for Deep Metric Learning","code":"https://github.com/billpsomas/Metrix_ICLR22","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":10,"model":"ResNet50 + S2SD","metrics":{"R@1":"89.5"},"uses_additional_data":true,"paper_date":"2020-09-17","paper":"/paper/s2sd-simultaneous-similarity-based-self","paper_url":"https://arxiv.org/abs/2009.08348v3","paper_title":"S2SD: Simultaneous Similarity-based Self-Distillation for Deep Metric Learning","code":"https://github.com/MLforHealth/S2SD","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":3,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":11,"model":"Recall@k Surrogate loss (ViT-B/16)","metrics":{"R@1":"89.5"},"uses_additional_data":false,"paper_date":"2021-08-25","paper":"/paper/recall-k-surrogate-loss-with-large-batches","paper_url":"https://arxiv.org/abs/2108.11179v2","paper_title":"Recall@k Surrogate Loss with Large Batches and Similarity Mixup","code":"https://github.com/yash0307/RecallatK_surrogate","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":11,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":12,"model":"ResNet-50 + Cross-Entropy","metrics":{"R@1":"89.3"},"uses_additional_data":true,"paper_date":"2020-03-19","paper":"/paper/metric-learning-cross-entropy-vs-pairwise","paper_url":"https://arxiv.org/abs/2003.08983v3","paper_title":"A unifying mutual information view of metric learning: cross-entropy vs. pairwise losses","code":"https://github.com/jeromerony/dml_cross_entropy","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":13,"model":"Hyp-DINO","metrics":{"R@1":"89.2"},"uses_additional_data":false,"paper_date":"2022-03-21","paper":"/paper/hyperbolic-vision-transformers-combining","paper_url":"https://arxiv.org/abs/2203.10833v2","paper_title":"Hyperbolic Vision Transformers: Combining Improvements in Metric Learning","code":"https://github.com/OML-Team/open-metric-learning","n_code_links":2,"syntology":{"n_ran":4,"n_unverified":2,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":14,"model":"ResNet50 + NIR","metrics":{"R@1":"89.1"},"uses_additional_data":true,"paper_date":"2022-03-16","paper":"/paper/non-isotropy-regularization-for-proxy-based","paper_url":"https://arxiv.org/abs/2203.08547v1","paper_title":"Non-isotropy Regularization for Proxy-based Deep Metric Learning","code":"https://github.com/explainableml/nonisotropicproxydml","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":1,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":15,"model":"Margin + DAS","metrics":{"R@1":"88.34"},"uses_additional_data":false,"paper_date":"2022-07-30","paper":"/paper/das-densely-anchored-sampling-for-deep-metric","paper_url":"https://arxiv.org/abs/2208.00119v1","paper_title":"DAS: Densely-Anchored Sampling for Deep Metric Learning","code":"https://github.com/lizhaoliu-Lec/DAS","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":16,"model":"BN-Inception + Proxy-Anchor","metrics":{"R@1":"88.3"},"uses_additional_data":true,"paper_date":"2020-03-31","paper":"/paper/proxy-anchor-loss-for-deep-metric-learning","paper_url":"https://arxiv.org/abs/2003.13911v1","paper_title":"Proxy Anchor Loss for Deep Metric Learning","code":"https://github.com/tjddus9597/Proxy-Anchor-CVPR2020","n_code_links":3,"syntology":null},{"rank_in_archive_order":17,"model":"Recall@k Surrogate loss (ResNet-50)","metrics":{"R@1":"88.3"},"uses_additional_data":false,"paper_date":"2021-08-25","paper":"/paper/recall-k-surrogate-loss-with-large-batches","paper_url":"https://arxiv.org/abs/2108.11179v2","paper_title":"Recall@k Surrogate Loss with Large Batches and Similarity Mixup","code":"https://github.com/yash0307/RecallatK_surrogate","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":11,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":18,"model":"ResNet-50 + Intra-Batch","metrics":{"R@1":"88.1"},"uses_additional_data":false,"paper_date":"2021-02-15","paper":"/paper/learning-intra-batch-connections-for-deep","paper_url":"https://arxiv.org/abs/2102.07753v3","paper_title":"Learning Intra-Batch Connections for Deep Metric Learning","code":"https://github.com/dvl-tum/intra_batch","n_code_links":2,"syntology":null},{"rank_in_archive_order":19,"model":"ABE + HORDE","metrics":{"R@1":"88.0"},"uses_additional_data":true,"paper_date":"2019-08-07","paper":"/paper/metric-learning-with-horde-high-order","paper_url":"https://arxiv.org/abs/1908.02735v1","paper_title":"Metric Learning With HORDE: High-Order Regularizer for Deep Embeddings","code":"https://github.com/pierre-jacob/ICCV2019-Horde","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":8,"n_samples":8,"n_pointer_only_licence":0}},{"rank_in_archive_order":20,"model":"MS + SEC + DAS","metrics":{"R@1":"87.8"},"uses_additional_data":false,"paper_date":"2022-07-30","paper":"/paper/das-densely-anchored-sampling-for-deep-metric","paper_url":"https://arxiv.org/abs/2208.00119v1","paper_title":"DAS: Densely-Anchored Sampling for Deep Metric Learning","code":"https://github.com/lizhaoliu-Lec/DAS","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":21,"model":"ResNet50 + DiVA","metrics":{"R@1":"87.6"},"uses_additional_data":true,"paper_date":"2020-04-28","paper":"/paper/diva-diverse-visual-feature-aggregation","paper_url":"https://arxiv.org/abs/2004.13458v4","paper_title":"DiVA: Diverse Visual Feature Aggregation for Deep Metric Learning","code":"https://github.com/Confusezius/ECCV2020_DiVA_MultiFeature_DML","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":2,"n_samples":5,"n_pointer_only_licence":3}},{"rank_in_archive_order":22,"model":"ProxyAnchor + DIML","metrics":{"R@1":"87.01"},"uses_additional_data":true,"paper_date":"2021-08-12","paper":"/paper/towards-interpretable-deep-metric-learning","paper_url":"https://arxiv.org/abs/2108.05889v1","paper_title":"Towards Interpretable Deep Metric Learning with Structural Matching","code":"https://github.com/wl-zhao/diml","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":2}},{"rank_in_archive_order":23,"model":"ResNet-50 + ProxyNCA++","metrics":{"R@1":"86.5"},"uses_additional_data":true,"paper_date":"2020-04-02","paper":"/paper/proxynca-revisiting-and-revitalizing-proxy","paper_url":"https://arxiv.org/abs/2004.01113v2","paper_title":"ProxyNCA++: Revisiting and Revitalizing Proxy Neighborhood Component Analysis","code":"https://github.com/euwern/proxynca_pp","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":1,"n_samples":5,"n_pointer_only_licence":0}},{"rank_in_archive_order":24,"model":"Gradient Surgery","metrics":{"R@1":"86.5"},"uses_additional_data":false,"paper_date":"2022-01-27","paper":"/paper/dissecting-the-impact-of-different-loss","paper_url":"https://arxiv.org/abs/2201.11307v1","paper_title":"Dissecting the impact of different loss functions with gradient surgery","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":25,"model":"Hyp-ViT","metrics":{"R@1":"86.5"},"uses_additional_data":true,"paper_date":"2022-03-21","paper":"/paper/hyperbolic-vision-transformers-combining","paper_url":"https://arxiv.org/abs/2203.10833v2","paper_title":"Hyperbolic Vision Transformers: Combining Improvements in Metric Learning","code":"https://github.com/OML-Team/open-metric-learning","n_code_links":2,"syntology":{"n_ran":4,"n_unverified":2,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":26,"model":"Group Loss","metrics":{"R@1":"85.6"},"uses_additional_data":false,"paper_date":"2019-12-01","paper":"/paper/the-group-loss-for-deep-metric-learning","paper_url":"https://arxiv.org/abs/1912.00385v4","paper_title":"The Group Loss for Deep Metric Learning","code":"https://github.com/dvl-tum/group_loss","n_code_links":2,"syntology":{"n_ran":12,"n_unverified":6,"n_samples":18,"n_pointer_only_licence":0}},{"rank_in_archive_order":27,"model":"ABE-8-512","metrics":{"R@1":"85.2"},"uses_additional_data":true,"paper_date":"2018-04-02","paper":"/paper/attention-based-ensemble-for-deep-metric","paper_url":"http://arxiv.org/abs/1804.00382v2","paper_title":"Attention-based Ensemble for Deep Metric Learning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":28,"model":"BN-Inception + SoftTriple","metrics":{"R@1":"84.5"},"uses_additional_data":true,"paper_date":"2019-09-11","paper":"/paper/softtriple-loss-deep-metric-learning-without","paper_url":"https://arxiv.org/abs/1909.05235v2","paper_title":"SoftTriple Loss: Deep Metric Learning Without Triplet Sampling","code":"https://github.com/idstcv/SoftTriple","n_code_links":5,"syntology":{"n_ran":2,"n_unverified":2,"n_samples":4,"n_pointer_only_licence":2}},{"rank_in_archive_order":29,"model":"ResNet50 (128) + PADS","metrics":{"R@1":"83.5"},"uses_additional_data":true,"paper_date":"2020-03-24","paper":"/paper/pads-policy-adapted-sampling-for-visual","paper_url":"https://arxiv.org/abs/2003.11113v2","paper_title":"PADS: Policy-Adapted Sampling for Visual Similarity Learning","code":"https://github.com/Confusezius/CVPR2020_PADS","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":6,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":30,"model":"CircleLoss","metrics":{"R@1":"83.4"},"uses_additional_data":false,"paper_date":"2020-02-25","paper":"/paper/circle-loss-a-unified-perspective-of-pair","paper_url":"https://arxiv.org/abs/2002.10857v2","paper_title":"Circle Loss: A Unified Perspective of Pair Similarity Optimization","code":"https://github.com/layumi/Person_reID_baseline_pytorch","n_code_links":16,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":1}},{"rank_in_archive_order":31,"model":"EPSHN(512)","metrics":{"R@1":"82.7"},"uses_additional_data":false,"paper_date":"2019-04-08","paper":"/paper/improved-embeddings-with-easy-positive","paper_url":"https://arxiv.org/abs/1904.04370v2","paper_title":"Improved Embeddings with Easy Positive Triplet Mining","code":"https://github.com/littleredxh/DREML","n_code_links":3,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":32,"model":"ResNet50 (128) + MIC","metrics":{"R@1":"82.6"},"uses_additional_data":true,"paper_date":"2019-09-25","paper":"/paper/mic-mining-interclass-characteristics-for","paper_url":"https://arxiv.org/abs/1909.11574v1","paper_title":"MIC: Mining Interclass Characteristics for Improved Metric Learning","code":"https://github.com/Confusezius/metric-learning-mining-interclass-characteristics","n_code_links":2,"syntology":null},{"rank_in_archive_order":33,"model":"ResNet-50 + Margin","metrics":{"R@1":"79.6"},"uses_additional_data":true,"paper_date":"2017-06-23","paper":"/paper/sampling-matters-in-deep-embedding-learning","paper_url":"http://arxiv.org/abs/1706.07567v2","paper_title":"Sampling Matters in Deep Embedding Learning","code":"https://github.com/CompVis/metric-learning-divide-and-conquer","n_code_links":6,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":34,"model":"GoogLeNet + HDML","metrics":{"R@1":"79.1"},"uses_additional_data":true,"paper_date":"2019-03-13","paper":"/paper/hardness-aware-deep-metric-learning","paper_url":"https://arxiv.org/abs/1903.05503v2","paper_title":"Hardness-Aware Deep Metric Learning","code":"https://github.com/wzzheng/HDML","n_code_links":2,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":35,"model":"EPSHN(64)","metrics":{"R@1":"75.5"},"uses_additional_data":false,"paper_date":"2019-04-08","paper":"/paper/improved-embeddings-with-easy-positive","paper_url":"https://arxiv.org/abs/1904.04370v2","paper_title":"Improved Embeddings with Easy Positive Triplet Mining","code":"https://github.com/littleredxh/DREML","n_code_links":3,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":36,"model":"SCT(64)","metrics":{"R@1":"73.2"},"uses_additional_data":false,"paper_date":"2020-07-24","paper":"/paper/hard-negative-examples-are-hard-but-useful","paper_url":"https://arxiv.org/abs/2007.12749v2","paper_title":"Hard negative examples are hard, but useful","code":"https://github.com/littleredxh/HardNegative","n_code_links":1,"syntology":null}],"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 6,885 of the 9,623 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":9623,"papers_checked":6885,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":2737},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-25T09:33:49+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":27,"rows_with_any_sample_ran":21,"distinct_papers_with_graph_line":22,"distinct_papers_with_any_sample_ran":18,"samples_over_distinct_papers":{"n_ran":56,"n_unverified":50,"n_samples":106,"n_pointer_only_licence":15,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":65,"n_unverified":66,"n_samples":131,"n_pointer_only_licence":16,"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"}}}