{"url":"/sota/image-generation-on-stl-10","task":{"name":"Image Generation","url":"/task/image-generation","note":null},"dataset":{"name":"STL-10","url":"/dataset/stl-10"},"category":"Computer Vision","categories":["Computer Vision","Medical","Miscellaneous","Natural Language Processing"],"category_note":null,"description":"**Image Generation** (synthesis) is the task of generating new images from an existing dataset.\r\n\r\n- **Unconditional generation** refers to generating samples unconditionally from the dataset, i.e. $p(y)$\r\n- **[Conditional image generation](/task/conditional-image-generation)** (subtask) refers to generating samples conditionally from the dataset, based on a label, i.e. $p(y|x)$.\r\n\r\nIn this section, you can find state-of-the-art leaderboards for **unconditional generation**. For conditional  generation, and other types of image generations, refer to the subtasks.\r\n\r\n<span style=\"color:grey; opacity: 0.6\">( Image credit: [StyleGAN](https://github.com/NVlabs/stylegan) )</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":["FID","Inception score","Model Size (MB)","Recall","NFE"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"FID":"lower","Inception score":"higher","Model Size (MB)":null,"Recall":"higher","NFE":null}},"counts":{"rows":31,"rows_with_code":28,"rows_with_paper_page":31,"rows_dated":31,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"Diffusion ProjectedGAN","metrics":{"FID":"6.91"},"uses_additional_data":false,"paper_date":"2022-06-05","paper":"/paper/diffusion-gan-training-gans-with-diffusion","paper_url":"https://arxiv.org/abs/2206.02262v4","paper_title":"Diffusion-GAN: Training GANs with Diffusion","code":"https://github.com/Zhendong-Wang/Diffusion-GAN","n_code_links":4,"syntology":{"n_ran":14,"n_unverified":4,"n_samples":18,"n_pointer_only_licence":0}},{"rank_in_archive_order":2,"model":"UNCSN++ (RVE) + ST","metrics":{"FID":"7.71","Inception score":"13.43"},"uses_additional_data":false,"paper_date":"2021-06-10","paper":"/paper/score-matching-model-for-unbounded-data-score-1","paper_url":"https://arxiv.org/abs/2106.05527v5","paper_title":"Soft Truncation: A Universal Training Technique of Score-based Diffusion Model for High Precision Score Estimation","code":"https://github.com/Kim-Dongjun/Soft-Truncation","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":8,"n_samples":12,"n_pointer_only_licence":0}},{"rank_in_archive_order":3,"model":"RDUOT","metrics":{"FID":"11.5","Recall":"0.49"},"uses_additional_data":false,"paper_date":"2023-11-28","paper":"/paper/robust-diffusion-gan-using-semi-unbalanced","paper_url":"https://arxiv.org/abs/2311.17101v2","paper_title":"A High-Quality Robust Diffusion Framework for Corrupted Dataset","code":"https://github.com/VinAIResearch/RDUOT","n_code_links":1,"syntology":{"n_ran":5,"n_unverified":7,"n_samples":12,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"Diffusion StyleGAN2","metrics":{"FID":"11.53"},"uses_additional_data":false,"paper_date":"2022-06-05","paper":"/paper/diffusion-gan-training-gans-with-diffusion","paper_url":"https://arxiv.org/abs/2206.02262v4","paper_title":"Diffusion-GAN: Training GANs with Diffusion","code":"https://github.com/Zhendong-Wang/Diffusion-GAN","n_code_links":4,"syntology":{"n_ran":14,"n_unverified":4,"n_samples":18,"n_pointer_only_licence":0}},{"rank_in_archive_order":5,"model":"MMD-PMish-NAS","metrics":{"FID":"11.61","Inception score":"11.79","Model Size (MB)":"19.47"},"uses_additional_data":false,"paper_date":"2024-10-23","paper":"/paper/enhancing-gans-with-mmd-neural-architecture","paper_url":"https://ieeexplore.ieee.org/document/10732016","paper_title":"Enhancing GANs with MMD Neural Architecture Search, PMish Activation Function, and Adaptive Rank Decomposition","code":"https://github.com/PrasannaPulakurthi/MMD-PMish-NAS","n_code_links":1,"syntology":null},{"rank_in_archive_order":6,"model":"MMD-AdversarialNAS","metrics":{"FID":"12.91","Inception score":"11.6","Model Size (MB)":"19.47"},"uses_additional_data":false,"paper_date":"2024-03-18","paper":"/paper/enhancing-gan-performance-through-neural","paper_url":"https://ieeexplore.ieee.org/document/10446488","paper_title":"Enhancing GAN Performance through Neural Architecture Search and Tensor Decomposition","code":"https://github.com/PrasannaPulakurthi/MMD-AdversarialNAS","n_code_links":1,"syntology":null},{"rank_in_archive_order":7,"model":"WaveDiff","metrics":{"FID":"12.93","NFE":"4","Recall":"0.41"},"uses_additional_data":false,"paper_date":"2022-11-29","paper":"/paper/wavelet-diffusion-models-are-fast-and","paper_url":"https://arxiv.org/abs/2211.16152v2","paper_title":"Wavelet Diffusion Models are fast and scalable Image Generators","code":"https://github.com/vinairesearch/wavediff","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"MMD-AdversarialNAS (Compressed Large)","metrics":{"FID":"13.06","Inception score":"11.28","Model Size (MB)":"5.35"},"uses_additional_data":false,"paper_date":"2024-03-18","paper":"/paper/enhancing-gan-performance-through-neural","paper_url":"https://ieeexplore.ieee.org/document/10446488","paper_title":"Enhancing GAN Performance through Neural Architecture Search and Tensor Decomposition","code":"https://github.com/PrasannaPulakurthi/MMD-AdversarialNAS","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"MMD-PMish-NAS (Compressed)","metrics":{"FID":"13.07","Inception score":"11.85","Model Size (MB)":"2.78"},"uses_additional_data":false,"paper_date":"2024-10-23","paper":"/paper/enhancing-gans-with-mmd-neural-architecture","paper_url":"https://ieeexplore.ieee.org/document/10732016","paper_title":"Enhancing GANs with MMD Neural Architecture Search, PMish Activation Function, and Adaptive Rank Decomposition","code":"https://github.com/PrasannaPulakurthi/MMD-PMish-NAS","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"RDGAN","metrics":{"FID":"13.07","Recall":"0.47"},"uses_additional_data":false,"paper_date":"2023-11-28","paper":"/paper/robust-diffusion-gan-using-semi-unbalanced","paper_url":"https://arxiv.org/abs/2311.17101v2","paper_title":"A High-Quality Robust Diffusion Framework for Corrupted Dataset","code":"https://github.com/VinAIResearch/RDUOT","n_code_links":1,"syntology":{"n_ran":5,"n_unverified":7,"n_samples":12,"n_pointer_only_licence":0}},{"rank_in_archive_order":11,"model":"MMD-AdversarialNAS (Compressed Small)","metrics":{"FID":"14.84","Inception score":"11.66","Model Size (MB)":"1.71"},"uses_additional_data":false,"paper_date":"2024-03-18","paper":"/paper/enhancing-gan-performance-through-neural","paper_url":"https://ieeexplore.ieee.org/document/10446488","paper_title":"Enhancing GAN Performance through Neural Architecture Search and Tensor Decomposition","code":"https://github.com/PrasannaPulakurthi/MMD-AdversarialNAS","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"Styleformer","metrics":{"FID":"15.17","Inception score":"11.01"},"uses_additional_data":false,"paper_date":"2021-06-13","paper":"/paper/styleformer-transformer-based-generative","paper_url":"https://arxiv.org/abs/2106.07023v3","paper_title":"Styleformer: Transformer based Generative Adversarial Networks with Style Vector","code":"https://github.com/BR-IDL/PaddleViT/blob/main/gan/Styleformer","n_code_links":3,"syntology":{"n_ran":4,"n_unverified":0,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":13,"model":"SNGAN-DCD (Latent)","metrics":{"FID":"17.68","Inception score":"9.33"},"uses_additional_data":false,"paper_date":"2020-04-05","paper":"/paper/discriminator-contrastive-divergence-semi","paper_url":"https://arxiv.org/abs/2004.01704v1","paper_title":"Discriminator Contrastive Divergence: Semi-Amortized Generative Modeling by Exploring Energy of the Discriminator","code":"https://github.com/MinkaiXu/Discriminator-Contrastive-Divergence","n_code_links":1,"syntology":null},{"rank_in_archive_order":14,"model":"TransGAN","metrics":{"FID":"18.28","Inception score":"10.43"},"uses_additional_data":false,"paper_date":"2021-02-14","paper":"/paper/transgan-two-transformers-can-make-one-strong","paper_url":"https://arxiv.org/abs/2102.07074v4","paper_title":"TransGAN: Two Pure Transformers Can Make One Strong GAN, and That Can Scale Up","code":"https://github.com/VITA-Group/TransGAN","n_code_links":10,"syntology":{"n_ran":8,"n_unverified":0,"n_samples":8,"n_pointer_only_licence":7}},{"rank_in_archive_order":15,"model":"PGMGAN","metrics":{"FID":"19.52","Inception score":"11.16"},"uses_additional_data":false,"paper_date":"2021-04-02","paper":"/paper/partition-guided-gans","paper_url":"https://arxiv.org/abs/2104.00816v2","paper_title":"Partition-Guided GANs","code":"https://github.com/alisadeghian/PGMGAN","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":1,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":16,"model":"EAGAN (G+D)","metrics":{"FID":"22.18","Inception score":"10.44"},"uses_additional_data":false,"paper_date":"2021-11-30","paper":"/paper/eagan-efficient-two-stage-evolutionary","paper_url":"https://arxiv.org/abs/2111.15097v2","paper_title":"EAGAN: Efficient Two-stage Evolutionary Architecture Search for GANs","code":"https://github.com/marsggbo/EAGAN","n_code_links":1,"syntology":null},{"rank_in_archive_order":17,"model":"SNGAN-DCD (Pixel)","metrics":{"FID":"22.25","Inception score":"9.25"},"uses_additional_data":false,"paper_date":"2020-04-05","paper":"/paper/discriminator-contrastive-divergence-semi","paper_url":"https://arxiv.org/abs/2004.01704v1","paper_title":"Discriminator Contrastive Divergence: Semi-Amortized Generative Modeling by Exploring Energy of the Discriminator","code":"https://github.com/MinkaiXu/Discriminator-Contrastive-Divergence","n_code_links":1,"syntology":null},{"rank_in_archive_order":18,"model":"EAGAN (G)","metrics":{"FID":"23.34","Inception score":"10.02"},"uses_additional_data":false,"paper_date":"2021-11-30","paper":"/paper/eagan-efficient-two-stage-evolutionary","paper_url":"https://arxiv.org/abs/2111.15097v2","paper_title":"EAGAN: Efficient Two-stage Evolutionary Architecture Search for GANs","code":"https://github.com/marsggbo/EAGAN","n_code_links":1,"syntology":null},{"rank_in_archive_order":19,"model":"E2GAN","metrics":{"FID":"25.35","Inception score":"9.51"},"uses_additional_data":false,"paper_date":"2020-07-17","paper":"/paper/off-policy-reinforcement-learning-for","paper_url":"https://arxiv.org/abs/2007.09180v1","paper_title":"Off-Policy Reinforcement Learning for Efficient and Effective GAN Architecture Search","code":"https://github.com/Yuantian013/E2GAN","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":4,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":20,"model":"aw-AutoGAN","metrics":{"FID":"26.32","Inception score":"9.59"},"uses_additional_data":false,"paper_date":"2020-12-05","paper":"/paper/adaptive-weighted-discriminator-for-training","paper_url":"https://arxiv.org/abs/2012.03149v2","paper_title":"Adaptive Weighted Discriminator for Training Generative Adversarial Networks","code":"https://github.com/vasily789/adaptive-weighted-gans","n_code_links":3,"syntology":null},{"rank_in_archive_order":21,"model":"DEGAS","metrics":{"FID":"28.76","Inception score":"9.71"},"uses_additional_data":false,"paper_date":"2019-12-02","paper":"/paper/degas-differentiable-efficient-generator","paper_url":"https://arxiv.org/abs/1912.00606v3","paper_title":"DEGAS: Differentiable Efficient Generator Search","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":22,"model":"AutoGAN","metrics":{"FID":"31.01","Inception score":"9.16"},"uses_additional_data":false,"paper_date":"2019-08-11","paper":"/paper/autogan-neural-architecture-search-for","paper_url":"https://arxiv.org/abs/1908.03835v1","paper_title":"AutoGAN: Neural Architecture Search for Generative Adversarial Networks","code":"https://github.com/TAMU-VITA/AutoGAN","n_code_links":2,"syntology":{"n_ran":1,"n_unverified":4,"n_samples":5,"n_pointer_only_licence":0}},{"rank_in_archive_order":23,"model":"aw-SN-GAN","metrics":{"FID":"34.72","Inception score":"9.61"},"uses_additional_data":false,"paper_date":"2020-12-05","paper":"/paper/adaptive-weighted-discriminator-for-training","paper_url":"https://arxiv.org/abs/2012.03149v2","paper_title":"Adaptive Weighted Discriminator for Training Generative Adversarial Networks","code":"https://github.com/vasily789/adaptive-weighted-gans","n_code_links":3,"syntology":null},{"rank_in_archive_order":24,"model":"Dist-GAN","metrics":{"FID":"36.19"},"uses_additional_data":false,"paper_date":"2018-03-23","paper":"/paper/dist-gan-an-improved-gan-using-distance","paper_url":"http://arxiv.org/abs/1803.08887v3","paper_title":"Dist-GAN: An Improved GAN using Distance Constraints","code":"https://github.com/tntrung/gan","n_code_links":1,"syntology":null},{"rank_in_archive_order":25,"model":"Improving MMD GAN","metrics":{"FID":"37.63","Inception score":"9.34"},"uses_additional_data":false,"paper_date":"2018-12-24","paper":"/paper/improving-mmd-gan-training-with-repulsive","paper_url":"http://arxiv.org/abs/1812.09916v4","paper_title":"Improving MMD-GAN Training with Repulsive Loss Function","code":"https://github.com/richardwth/MMD-GAN","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":9,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":26,"model":"SN-GAN","metrics":{"FID":"40.1","Inception score":"9.10"},"uses_additional_data":false,"paper_date":"2018-02-16","paper":"/paper/spectral-normalization-for-generative","paper_url":"http://arxiv.org/abs/1802.05957v1","paper_title":"Spectral Normalization for Generative Adversarial Networks","code":"https://github.com/pytorch/pytorch/blob/a77b391de723a69fb59ff6ae9d1236ca93f03a97/torch/nn/utils/spectral_norm.py","n_code_links":38,"syntology":{"n_ran":16,"n_unverified":15,"n_samples":31,"n_pointer_only_licence":14}},{"rank_in_archive_order":27,"model":"DC-VAE","metrics":{"FID":"41.9","Inception score":"8.1"},"uses_additional_data":false,"paper_date":"2020-11-19","paper":"/paper/dual-contradistinctive-generative-autoencoder-1","paper_url":"https://arxiv.org/abs/2011.10063v1","paper_title":"Dual Contradistinctive Generative Autoencoder","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":28,"model":"ProbGAN","metrics":{"FID":"46.74","Inception score":"8.87"},"uses_additional_data":false,"paper_date":"2019-05-01","paper":"/paper/probgan-towards-probabilistic-gan-with","paper_url":"https://openreview.net/forum?id=H1l7bnR5Ym","paper_title":"ProbGAN: Towards Probabilistic GAN with Theoretical Guarantees","code":"https://github.com/hehaodele/ProbGAN","n_code_links":1,"syntology":null},{"rank_in_archive_order":29,"model":"PeerGAN","metrics":{"FID":"51.37"},"uses_additional_data":false,"paper_date":"2021-01-19","paper":"/paper/peergan-generative-adversarial-networks-with","paper_url":"https://arxiv.org/abs/2101.07524v3","paper_title":"DuelGAN: A Duel Between Two Discriminators Stabilizes the GAN Training","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":30,"model":"QSNGAN","metrics":{"FID":"59.611","Inception score":"4.987"},"uses_additional_data":false,"paper_date":"2021-04-19","paper":"/paper/quaternion-generative-adversarial-networks","paper_url":"https://arxiv.org/abs/2104.09630v2","paper_title":"Quaternion Generative Adversarial Networks","code":"https://github.com/eleGAN23/QVAE","n_code_links":3,"syntology":null},{"rank_in_archive_order":31,"model":"D2GAN","metrics":{"Inception score":"7.98"},"uses_additional_data":false,"paper_date":"2017-09-12","paper":"/paper/dual-discriminator-generative-adversarial","paper_url":"http://arxiv.org/abs/1709.03831v1","paper_title":"Dual Discriminator Generative Adversarial Nets","code":"https://github.com/tund/D2GAN","n_code_links":2,"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":12,"rows_with_any_sample_ran":11,"distinct_papers_with_graph_line":10,"distinct_papers_with_any_sample_ran":9,"samples_over_distinct_papers":{"n_ran":56,"n_unverified":52,"n_samples":108,"n_pointer_only_licence":21,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":75,"n_unverified":63,"n_samples":138,"n_pointer_only_licence":21,"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"}}}