{"url":"/sota/metric-learning-on-in-shop-1","task":{"name":"Metric Learning","url":"/task/metric-learning","note":null},"dataset":{"name":"In-Shop","url":"/dataset/in-shop"},"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":15,"rows_with_code":11,"rows_with_paper_page":15,"rows_dated":15,"rows_using_additional_data":7},"rows":[{"rank_in_archive_order":1,"model":"Unicom+ViT-L@336px","metrics":{"R@1":"96.7"},"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":"STIR","metrics":{"R@1":"95"},"uses_additional_data":false,"paper_date":"2023-04-26","paper":"/paper/stir-siamese-transformer-for-image-retrieval","paper_url":"https://arxiv.org/abs/2304.13393v2","paper_title":"STIR: Siamese Transformer for Image Retrieval Postprocessing","code":"https://github.com/OML-Team/open-metric-learning/tree/main/pipelines/postprocessing/pairwise_postprocessing","n_code_links":1,"syntology":null},{"rank_in_archive_order":3,"model":"MGA","metrics":{"R@1":"94.3"},"uses_additional_data":false,"paper_date":"2023-02-16","paper":"/paper/fashion-image-retrieval-with-multi-granular-1","paper_url":"https://arxiv.org/abs/2302.08902v4","paper_title":"Fashion Image Retrieval with Multi-Granular Alignment","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":4,"model":"Hyp-ViT","metrics":{"R@1":"92.5"},"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":5,"model":"Hyp-DINO","metrics":{"R@1":"92.4"},"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":6,"model":"CCL (ResNet-50)","metrics":{"R@1":"92.31"},"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":7,"model":"Gradient Surgery","metrics":{"R@1":"92.21"},"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":8,"model":"ResNet-50 + Metrix","metrics":{"R@1":"92.2"},"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":9,"model":"ViT-Triplet","metrics":{"R@1":"92.1"},"uses_additional_data":false,"paper_date":"2023-04-26","paper":"/paper/stir-siamese-transformer-for-image-retrieval","paper_url":"https://arxiv.org/abs/2304.13393v2","paper_title":"STIR: Siamese Transformer for Image Retrieval Postprocessing","code":"https://github.com/OML-Team/open-metric-learning/tree/main/pipelines/postprocessing/pairwise_postprocessing","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"EfficientDML-VPTSP-G/512","metrics":{"R@1":"92.1"},"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":11,"model":"NED","metrics":{"R@1":"91.3"},"uses_additional_data":true,"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":12,"model":"ResNet-50 + ProxyNCA++","metrics":{"R@1":"90.9"},"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":13,"model":"ResNet-50 + Cross-Entropy","metrics":{"R@1":"90.6"},"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":14,"model":"SCT(512)","metrics":{"R@1":"90"},"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},{"rank_in_archive_order":15,"model":"EPSHN(512)","metrics":{"R@1":"87.8"},"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}}],"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":8,"rows_with_any_sample_ran":8,"distinct_papers_with_graph_line":7,"distinct_papers_with_any_sample_ran":7,"samples_over_distinct_papers":{"n_ran":18,"n_unverified":8,"n_samples":26,"n_pointer_only_licence":7,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":22,"n_unverified":10,"n_samples":32,"n_pointer_only_licence":7,"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"}}}