{"url":"/sota/domain-generalization-on-domainnet","task":{"name":"Domain Generalization","url":"/task/domain-generalization","note":null},"dataset":{"name":"DomainNet","url":"/dataset/domainnet"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"The idea of **Domain Generalization** is to learn from one or multiple training domains, to extract a domain-agnostic model which can be applied to an unseen domain\n\n\n<span class=\"description-source\">Source: [Diagram Image Retrieval using Sketch-Based Deep Learning and Transfer Learning ](https://arxiv.org/abs/2004.10780)</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 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 Accuracy":"higher"}},"counts":{"rows":38,"rows_with_code":34,"rows_with_paper_page":38,"rows_dated":38,"rows_using_additional_data":6},"rows":[{"rank_in_archive_order":1,"model":"L2C (CLIP, ViT-L/14)","metrics":{"Average Accuracy":"67.4"},"uses_additional_data":false,"paper_date":"2025-01-15","paper":"/paper/learning-to-adapt-frozen-clip-for-few-shot","paper_url":"https://openreview.net/forum?id=TD3SGJfBC7","paper_title":"Learning to Adapt Frozen CLIP for Few-Shot Test-Time Domain Adaptation","code":"https://github.com/chi-chi-zx/L2C","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"PromptStyler (CLIP, ViT-L/14)","metrics":{"Average Accuracy":"65.5"},"uses_additional_data":false,"paper_date":"2023-07-27","paper":"/paper/promptstyler-prompt-driven-style-generation","paper_url":"https://arxiv.org/abs/2307.15199v2","paper_title":"PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization","code":"https://github.com/zhanghr2001/promptta","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":3,"model":"VDPG (CLIP, ViT-L/14)","metrics":{"Average Accuracy":"65.2"},"uses_additional_data":false,"paper_date":"2024-05-05","paper":"/paper/adapting-to-distribution-shift-by-visual","paper_url":"https://arxiv.org/abs/2405.02797v1","paper_title":"Adapting to Distribution Shift by Visual Domain Prompt Generation","code":"https://github.com/guliisgreat/vdpg","n_code_links":1,"syntology":{"n_ran":7,"n_unverified":0,"n_samples":7,"n_pointer_only_licence":7}},{"rank_in_archive_order":4,"model":"VL2V-SD (CLIP, ViT-B/16)","metrics":{"Average Accuracy":"62.79"},"uses_additional_data":false,"paper_date":"2023-10-12","paper":"/paper/distilling-from-vision-language-models-for","paper_url":"https://arxiv.org/abs/2310.08255v2","paper_title":"Leveraging Vision-Language Models for Improving Domain Generalization in Image Classification","code":"https://github.com/val-iisc/VL2V-ADiP","n_code_links":1,"syntology":{"n_ran":12,"n_unverified":3,"n_samples":15,"n_pointer_only_licence":0}},{"rank_in_archive_order":5,"model":"MoA (OpenCLIP, ViT-B/16)","metrics":{"Average Accuracy":"62.7"},"uses_additional_data":false,"paper_date":"2023-10-17","paper":"/paper/domain-generalization-using-large-pretrained","paper_url":"https://arxiv.org/abs/2310.11031v2","paper_title":"Domain Generalization Using Large Pretrained Models with Mixture-of-Adapters","code":"https://github.com/KU-CVLAB/MoA","n_code_links":1,"syntology":null},{"rank_in_archive_order":6,"model":"CAR-FT (CLIP, ViT-B/16)","metrics":{"Average Accuracy":"62.5"},"uses_additional_data":true,"paper_date":"2022-11-29","paper":"/paper/context-aware-robust-fine-tuning","paper_url":"https://arxiv.org/abs/2211.16175v1","paper_title":"Context-Aware Robust Fine-Tuning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":7,"model":"SIMPLE+","metrics":{"Average Accuracy":"61.9"},"uses_additional_data":false,"paper_date":"2023-05-01","paper":"/paper/simple-specialized-model-sample-matching-for","paper_url":"https://openreview.net/forum?id=BqrPeZ_e5P","paper_title":"SIMPLE: Specialized Model-Sample Matching for Domain Generalization","code":"https://github.com/microsoft/SeqML/tree/main/SIMPLE","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"GMDG (RegNetY-16GF, SWAD)","metrics":{"Average Accuracy":"61.3"},"uses_additional_data":false,"paper_date":"2024-02-29","paper":"/paper/rethinking-multi-domain-generalization-with-a","paper_url":"https://arxiv.org/abs/2402.18853v2","paper_title":"Rethinking Multi-domain Generalization with A General Learning Objective","code":"https://github.com/zhaorui-tan/gmdg","n_code_links":2,"syntology":null},{"rank_in_archive_order":9,"model":"L2C (CLIP, ViT-B/16)","metrics":{"Average Accuracy":"61.2"},"uses_additional_data":false,"paper_date":"2025-01-15","paper":"/paper/learning-to-adapt-frozen-clip-for-few-shot","paper_url":"https://openreview.net/forum?id=TD3SGJfBC7","paper_title":"Learning to Adapt Frozen CLIP for Few-Shot Test-Time Domain Adaptation","code":"https://github.com/chi-chi-zx/L2C","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"Ensemble of Averages (RegNetY-16GF)","metrics":{"Average Accuracy":"60.9"},"uses_additional_data":true,"paper_date":"2021-10-21","paper":"/paper/ensemble-of-averages-improving-model","paper_url":"https://arxiv.org/abs/2110.10832v4","paper_title":"Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization","code":"https://github.com/salesforce/ensemble-of-averages","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":2}},{"rank_in_archive_order":11,"model":"MIRO (RegNetY-16GF, SWAD)","metrics":{"Average Accuracy":"60.7"},"uses_additional_data":true,"paper_date":"2022-03-21","paper":"/paper/domain-generalization-by-mutual-information","paper_url":"https://arxiv.org/abs/2203.10789v2","paper_title":"Domain Generalization by Mutual-Information Regularization with Pre-trained Models","code":"https://github.com/kakaobrain/miro","n_code_links":1,"syntology":{"n_ran":11,"n_unverified":2,"n_samples":13,"n_pointer_only_licence":0}},{"rank_in_archive_order":12,"model":"SPG (CLIP, ViT-B/16)","metrics":{"Average Accuracy":"60.1"},"uses_additional_data":false,"paper_date":"2024-04-30","paper":"/paper/soft-prompt-generation-for-domain","paper_url":"https://arxiv.org/abs/2404.19286v2","paper_title":"Soft Prompt Generation for Domain Generalization","code":"https://github.com/renytek13/soft-prompt-generation-with-cgan","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":3,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":13,"model":"VDPG (CLIP, ViT-B/16)","metrics":{"Average Accuracy":"59.8"},"uses_additional_data":false,"paper_date":"2024-05-05","paper":"/paper/adapting-to-distribution-shift-by-visual","paper_url":"https://arxiv.org/abs/2405.02797v1","paper_title":"Adapting to Distribution Shift by Visual Domain Prompt Generation","code":"https://github.com/guliisgreat/vdpg","n_code_links":1,"syntology":{"n_ran":7,"n_unverified":0,"n_samples":7,"n_pointer_only_licence":7}},{"rank_in_archive_order":14,"model":"UniDG + CORAL + ConvNeXt-B","metrics":{"Average Accuracy":"59.5"},"uses_additional_data":false,"paper_date":"2023-10-16","paper":"/paper/towards-unified-and-effective-domain","paper_url":"https://arxiv.org/abs/2310.10008v1","paper_title":"Towards Unified and Effective Domain Generalization","code":"https://github.com/invictus717/UniDG","n_code_links":1,"syntology":null},{"rank_in_archive_order":15,"model":"PromptStyler (CLIP, ViT-B/16)","metrics":{"Average Accuracy":"59.4"},"uses_additional_data":false,"paper_date":"2023-07-27","paper":"/paper/promptstyler-prompt-driven-style-generation","paper_url":"https://arxiv.org/abs/2307.15199v2","paper_title":"PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization","code":"https://github.com/zhanghr2001/promptta","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":"SEDGE+","metrics":{"Average Accuracy":"54.7"},"uses_additional_data":false,"paper_date":"2022-03-09","paper":"/paper/domain-generalization-using-pretrained-models","paper_url":"https://arxiv.org/abs/2203.04600v1","paper_title":"Domain Generalization using Pretrained Models without Fine-tuning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":17,"model":"Ensemble of Averages (ResNeXt-50 32x4d)","metrics":{"Average Accuracy":"54.6"},"uses_additional_data":true,"paper_date":"2021-10-21","paper":"/paper/ensemble-of-averages-improving-model","paper_url":"https://arxiv.org/abs/2110.10832v4","paper_title":"Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization","code":"https://github.com/salesforce/ensemble-of-averages","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":2}},{"rank_in_archive_order":18,"model":"GMDG (RegNetY-16GF)","metrics":{"Average Accuracy":"54.6"},"uses_additional_data":false,"paper_date":"2024-02-29","paper":"/paper/rethinking-multi-domain-generalization-with-a","paper_url":"https://arxiv.org/abs/2402.18853v2","paper_title":"Rethinking Multi-domain Generalization with A General Learning Objective","code":"https://github.com/zhaorui-tan/gmdg","n_code_links":2,"syntology":null},{"rank_in_archive_order":19,"model":"Hybrid-SF-MoE","metrics":{"Average Accuracy":"52.0"},"uses_additional_data":true,"paper_date":"2022-06-08","paper":"/paper/sparse-fusion-mixture-of-experts-are-domain","paper_url":"https://arxiv.org/abs/2206.04046v6","paper_title":"Sparse Mixture-of-Experts are Domain Generalizable Learners","code":"https://github.com/luodian/sf-moe-dg","n_code_links":2,"syntology":{"n_ran":7,"n_unverified":7,"n_samples":14,"n_pointer_only_licence":14}},{"rank_in_archive_order":20,"model":"CADG","metrics":{"Average Accuracy":"51.6"},"uses_additional_data":true,"paper_date":"2022-03-31","paper":"/paper/cadg-a-model-based-on-cross-attention-for","paper_url":"https://arxiv.org/abs/2203.17067v3","paper_title":"CADG: A Model Based on Cross Attention for Domain Generalization","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":21,"model":"SPG (CLIP, ResNet-50)","metrics":{"Average Accuracy":"50.1"},"uses_additional_data":false,"paper_date":"2024-04-30","paper":"/paper/soft-prompt-generation-for-domain","paper_url":"https://arxiv.org/abs/2404.19286v2","paper_title":"Soft Prompt Generation for Domain Generalization","code":"https://github.com/renytek13/soft-prompt-generation-with-cgan","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":3,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":22,"model":"PromptStyler (CLIP, ResNet-50)","metrics":{"Average Accuracy":"49.5"},"uses_additional_data":false,"paper_date":"2023-07-27","paper":"/paper/promptstyler-prompt-driven-style-generation","paper_url":"https://arxiv.org/abs/2307.15199v2","paper_title":"PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization","code":"https://github.com/zhanghr2001/promptta","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":23,"model":"SIMPLE","metrics":{"Average Accuracy":"49.2"},"uses_additional_data":false,"paper_date":"2023-05-01","paper":"/paper/simple-specialized-model-sample-matching-for","paper_url":"https://openreview.net/forum?id=BqrPeZ_e5P","paper_title":"SIMPLE: Specialized Model-Sample Matching for Domain Generalization","code":"https://github.com/microsoft/SeqML/tree/main/SIMPLE","n_code_links":1,"syntology":null},{"rank_in_archive_order":24,"model":"GMoE-S/16","metrics":{"Average Accuracy":"48.7"},"uses_additional_data":false,"paper_date":"2022-06-08","paper":"/paper/sparse-fusion-mixture-of-experts-are-domain","paper_url":"https://arxiv.org/abs/2206.04046v6","paper_title":"Sparse Mixture-of-Experts are Domain Generalizable Learners","code":"https://github.com/luodian/sf-moe-dg","n_code_links":2,"syntology":{"n_ran":7,"n_unverified":7,"n_samples":14,"n_pointer_only_licence":14}},{"rank_in_archive_order":25,"model":"Ensemble of Averages (ResNet-50)","metrics":{"Average Accuracy":"47.4"},"uses_additional_data":false,"paper_date":"2021-10-21","paper":"/paper/ensemble-of-averages-improving-model","paper_url":"https://arxiv.org/abs/2110.10832v4","paper_title":"Ensemble of Averages: Improving Model Selection and Boosting Performance in Domain Generalization","code":"https://github.com/salesforce/ensemble-of-averages","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":2}},{"rank_in_archive_order":26,"model":"GMDG (ResNet-50, SWAD)","metrics":{"Average Accuracy":"47.3"},"uses_additional_data":false,"paper_date":"2024-02-29","paper":"/paper/rethinking-multi-domain-generalization-with-a","paper_url":"https://arxiv.org/abs/2402.18853v2","paper_title":"Rethinking Multi-domain Generalization with A General Learning Objective","code":"https://github.com/zhaorui-tan/gmdg","n_code_links":2,"syntology":null},{"rank_in_archive_order":27,"model":"MIRO (ResNet-50, SWAD)","metrics":{"Average Accuracy":"47.0"},"uses_additional_data":false,"paper_date":"2022-03-21","paper":"/paper/domain-generalization-by-mutual-information","paper_url":"https://arxiv.org/abs/2203.10789v2","paper_title":"Domain Generalization by Mutual-Information Regularization with Pre-trained Models","code":"https://github.com/kakaobrain/miro","n_code_links":1,"syntology":{"n_ran":11,"n_unverified":2,"n_samples":13,"n_pointer_only_licence":0}},{"rank_in_archive_order":28,"model":"DDG","metrics":{"Average Accuracy":"46.93"},"uses_additional_data":false,"paper_date":"2022-05-27","paper":"/paper/dynamic-domain-generalization","paper_url":"https://arxiv.org/abs/2205.13913v1","paper_title":"Dynamic Domain Generalization","code":"https://github.com/metavisionlab/ddg","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":29,"model":"AdaClust (ResNet-50, SWAD)","metrics":{"Average Accuracy":"46.7"},"uses_additional_data":false,"paper_date":"2021-12-09","paper":"/paper/adaptive-methods-for-aggregated-domain","paper_url":"https://arxiv.org/abs/2112.04766v2","paper_title":"Adaptive Methods for Aggregated Domain Generalization","code":"https://github.com/xavierohan/AdaClust_DomainBed","n_code_links":1,"syntology":null},{"rank_in_archive_order":30,"model":"SWAD (ResNet-50)","metrics":{"Average Accuracy":"46.5"},"uses_additional_data":false,"paper_date":"2021-02-17","paper":"/paper/domain-generalization-needs-stochastic-weight","paper_url":"https://arxiv.org/abs/2102.08604v4","paper_title":"SWAD: Domain Generalization by Seeking Flat Minima","code":"https://github.com/khanrc/swad","n_code_links":4,"syntology":{"n_ran":2,"n_unverified":1,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":31,"model":"SEDGE","metrics":{"Average Accuracy":"46.3"},"uses_additional_data":false,"paper_date":"2022-03-09","paper":"/paper/domain-generalization-using-pretrained-models","paper_url":"https://arxiv.org/abs/2203.04600v1","paper_title":"Domain Generalization using Pretrained Models without Fine-tuning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":32,"model":"GMDG (ResNet-50)","metrics":{"Average Accuracy":"44.6"},"uses_additional_data":false,"paper_date":"2024-02-29","paper":"/paper/rethinking-multi-domain-generalization-with-a","paper_url":"https://arxiv.org/abs/2402.18853v2","paper_title":"Rethinking Multi-domain Generalization with A General Learning Objective","code":"https://github.com/zhaorui-tan/gmdg","n_code_links":2,"syntology":null},{"rank_in_archive_order":33,"model":"Meta-DMoE (ResNet-50)","metrics":{"Average Accuracy":"44.2"},"uses_additional_data":false,"paper_date":"2022-10-08","paper":"/paper/meta-dmoe-adapting-to-domain-shift-by-meta","paper_url":"https://arxiv.org/abs/2210.03885v2","paper_title":"Meta-DMoE: Adapting to Domain Shift by Meta-Distillation from Mixture-of-Experts","code":"https://github.com/n3il666/meta-dmoe","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":7,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":34,"model":"POEM","metrics":{"Average Accuracy":"44.0"},"uses_additional_data":false,"paper_date":"2023-05-22","paper":"/paper/poem-polarization-of-embeddings-for-domain","paper_url":"https://arxiv.org/abs/2305.13046v1","paper_title":"POEM: Polarization of Embeddings for Domain-Invariant Representations","code":"https://github.com/josangyoung/official-poem","n_code_links":1,"syntology":{"n_ran":9,"n_unverified":3,"n_samples":12,"n_pointer_only_licence":0}},{"rank_in_archive_order":35,"model":"DMG (ResNet-50)","metrics":{"Average Accuracy":"43.63"},"uses_additional_data":false,"paper_date":"2020-08-28","paper":"/paper/learning-to-balance-specificity-and","paper_url":"https://arxiv.org/abs/2008.12839v1","paper_title":"Learning to Balance Specificity and Invariance for In and Out of Domain Generalization","code":"https://github.com/prithv1/DMG","n_code_links":1,"syntology":null},{"rank_in_archive_order":36,"model":"MetaReg (ResNet-50)","metrics":{"Average Accuracy":"43.62"},"uses_additional_data":false,"paper_date":"2020-08-28","paper":"/paper/learning-to-balance-specificity-and","paper_url":"https://arxiv.org/abs/2008.12839v1","paper_title":"Learning to Balance Specificity and Invariance for In and Out of Domain Generalization","code":"https://github.com/prithv1/DMG","n_code_links":1,"syntology":null},{"rank_in_archive_order":37,"model":"AdaClust (ResNet-50)","metrics":{"Average Accuracy":"43.3"},"uses_additional_data":false,"paper_date":"2021-12-09","paper":"/paper/adaptive-methods-for-aggregated-domain","paper_url":"https://arxiv.org/abs/2112.04766v2","paper_title":"Adaptive Methods for Aggregated Domain Generalization","code":"https://github.com/xavierohan/AdaClust_DomainBed","n_code_links":1,"syntology":null},{"rank_in_archive_order":38,"model":"Fishr (ResNet-50)","metrics":{"Average Accuracy":"41.8"},"uses_additional_data":false,"paper_date":"2021-09-07","paper":"/paper/fishr-invariant-gradient-variances-for-out-of","paper_url":"https://arxiv.org/abs/2109.02934v3","paper_title":"Fishr: Invariant Gradient Variances for Out-of-Distribution Generalization","code":"https://github.com/facebookresearch/DomainBed","n_code_links":2,"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 6,264 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":20,"rows_with_any_sample_ran":16,"distinct_papers_with_graph_line":12,"distinct_papers_with_any_sample_ran":10,"samples_over_distinct_papers":{"n_ran":58,"n_unverified":27,"n_samples":85,"n_pointer_only_licence":25,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":93,"n_unverified":41,"n_samples":134,"n_pointer_only_licence":50,"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"}}}