{"url":"/sota/weakly-supervised-semantic-segmentation-on-4","task":{"name":"Weakly-Supervised Semantic Segmentation","url":"/task/weakly-supervised-semantic-segmentation","note":null},"dataset":{"name":"COCO 2014 val","url":null},"category":"Computer Vision","categories":["Computer Code","Computer Vision","Medical","Robots"],"category_note":null,"description":"The semantic segmentation task is to assign a label from a label set to each pixel in an image. In the case of fully supervised setting, the dataset  consists of images and their corresponding\r\npixel-level class-specific annotations (expensive pixel-level annotations). However, in the\r\nweakly-supervised setting, the dataset consists of images and corresponding annotations that\r\nare relatively easy to obtain, such as tags/labels of objects present in the image.\r\n\r\n<span style=\"color:grey; opacity: 0.6\">( Image credit: [Weakly-Supervised Semantic Segmentation Network with Deep Seeded Region Growing](http://openaccess.thecvf.com/content_cvpr_2018/papers/Huang_Weakly-Supervised_Semantic_Segmentation_CVPR_2018_paper.pdf) )</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":["mIoU"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"mIoU":null}},"counts":{"rows":39,"rows_with_code":38,"rows_with_paper_page":39,"rows_dated":39,"rows_using_additional_data":6},"rows":[{"rank_in_archive_order":1,"model":"DHR (Swin-L, Mask2Former)","metrics":{"mIoU":"56.8"},"uses_additional_data":false,"paper_date":"2024-03-30","paper":"/paper/dhr-dual-features-driven-hierarchical","paper_url":"https://arxiv.org/abs/2404.00380v2","paper_title":"DHR: Dual Features-Driven Hierarchical Rebalancing in Inter- and Intra-Class Regions for Weakly-Supervised Semantic Segmentation","code":"https://github.com/shjo-april/DHR","n_code_links":1,"syntology":{"n_ran":9,"n_unverified":0,"n_samples":9,"n_pointer_only_licence":9}},{"rank_in_archive_order":2,"model":"SemPLeS (Swin-L)","metrics":{"mIoU":"56.1"},"uses_additional_data":false,"paper_date":"2024-01-22","paper":"/paper/semples-semantic-prompt-learning-for-weakly","paper_url":"https://arxiv.org/abs/2401.11791v4","paper_title":"Semantic Prompt Learning for Weakly-Supervised Semantic Segmentation","code":"https://github.com/nvlabs/semples","n_code_links":1,"syntology":null},{"rank_in_archive_order":3,"model":"WSSS-SAM(DeepLabV2-ResNet101)","metrics":{"mIoU":"55.6"},"uses_additional_data":true,"paper_date":"2023-05-02","paper":"/paper/an-alternative-to-wsss-an-empirical-study-of","paper_url":"https://arxiv.org/abs/2305.01586v2","paper_title":"An Alternative to WSSS? An Empirical Study of the Segment Anything Model (SAM) on Weakly-Supervised Semantic Segmentation Problems","code":"https://github.com/weixuansun/wsss_sam","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":5,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"FMA-WSSS (Swin-L)","metrics":{"mIoU":"55.4"},"uses_additional_data":false,"paper_date":"2023-12-06","paper":"/paper/foundation-model-assisted-weakly-supervised","paper_url":"https://arxiv.org/abs/2312.03585v2","paper_title":"Foundation Model Assisted Weakly Supervised Semantic Segmentation","code":"https://github.com/HAL-42/FMA-WSSS","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":1,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":5,"model":"CoSA (SWIN-B, multi-stage)","metrics":{"mIoU":"53.7"},"uses_additional_data":false,"paper_date":"2024-02-27","paper":"/paper/weakly-supervised-co-training-with-swapping","paper_url":"https://arxiv.org/abs/2402.17891v2","paper_title":"Weakly Supervised Co-training with Swapping Assignments for Semantic Segmentation","code":"https://github.com/youshyee/cosa","n_code_links":1,"syntology":null},{"rank_in_archive_order":6,"model":"CoSA (ViT-B, single-stage)","metrics":{"mIoU":"51.1"},"uses_additional_data":false,"paper_date":"2024-02-27","paper":"/paper/weakly-supervised-co-training-with-swapping","paper_url":"https://arxiv.org/abs/2402.17891v2","paper_title":"Weakly Supervised Co-training with Swapping Assignments for Semantic Segmentation","code":"https://github.com/youshyee/cosa","n_code_links":1,"syntology":null},{"rank_in_archive_order":7,"model":"WeakTr (ViT-S, multi-stage)","metrics":{"mIoU":"50.3"},"uses_additional_data":false,"paper_date":"2023-04-03","paper":"/paper/weaktr-exploring-plain-vision-transformer-for","paper_url":"https://arxiv.org/abs/2304.01184v2","paper_title":"WeakTr: Exploring Plain Vision Transformer for Weakly-supervised Semantic Segmentation","code":"https://github.com/hustvl/weaktr","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":8,"model":"MARS (ResNet-101, multi-stage)","metrics":{"mIoU":"49.4"},"uses_additional_data":false,"paper_date":"2023-04-19","paper":"/paper/mars-model-agnostic-biased-object-removal","paper_url":"https://arxiv.org/abs/2304.09913v1","paper_title":"MARS: Model-agnostic Biased Object Removal without Additional Supervision for Weakly-Supervised Semantic Segmentation","code":"https://github.com/shjo-april/mars","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":2,"n_samples":2,"n_pointer_only_licence":2}},{"rank_in_archive_order":9,"model":"WeakTr (DeiT-S, multi-stage)","metrics":{"mIoU":"46.9"},"uses_additional_data":false,"paper_date":"2023-04-03","paper":"/paper/weaktr-exploring-plain-vision-transformer-for","paper_url":"https://arxiv.org/abs/2304.01184v2","paper_title":"WeakTr: Exploring Plain Vision Transformer for Weakly-supervised Semantic Segmentation","code":"https://github.com/hustvl/weaktr","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":10,"model":"RS+EPM (ResNet-101, multi-stage)","metrics":{"mIoU":"46.4"},"uses_additional_data":false,"paper_date":"2022-04-14","paper":"/paper/recurseed-and-certainmix-for-weakly","paper_url":"https://arxiv.org/abs/2204.06754v4","paper_title":"RecurSeed and EdgePredictMix: Pseudo-Label Refinement Learning for Weakly Supervised Semantic Segmentation across Single- and Multi-Stage Frameworks","code":"https://github.com/shjo-april/recurseed_and_edgepredictmix","n_code_links":2,"syntology":null},{"rank_in_archive_order":11,"model":"T2MDiffusion(DeepLabV2-ResNet101)","metrics":{"mIoU":"45.7"},"uses_additional_data":true,"paper_date":"2023-09-08","paper":"/paper/from-text-to-mask-localizing-entities-using","paper_url":"https://arxiv.org/abs/2309.04109v2","paper_title":"From Text to Mask: Localizing Entities Using the Attention of Text-to-Image Diffusion Models","code":"https://github.com/Big-Brother-Pikachu/Text2Mask","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"FBR","metrics":{"mIoU":"45.6"},"uses_additional_data":false,"paper_date":"2024-06-22","paper":"/paper/fine-grained-background-representation-for","paper_url":"https://arxiv.org/abs/2406.15755v1","paper_title":"Fine-grained Background Representation for Weakly Supervised Semantic Segmentation","code":"https://github.com/YininKorea/FBR","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"CLIP-ES(DeepLabV2-ResNet101)","metrics":{"mIoU":"45.4"},"uses_additional_data":false,"paper_date":"2022-12-16","paper":"/paper/clip-is-also-an-efficient-segmenter-a-text","paper_url":"https://arxiv.org/abs/2212.09506v3","paper_title":"CLIP is Also an Efficient Segmenter: A Text-Driven Approach for Weakly Supervised Semantic Segmentation","code":"https://github.com/linyq2117/clip-es","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":14,"model":"ACR(DeeplabV1-ResNet38)","metrics":{"mIoU":"45.3"},"uses_additional_data":false,"paper_date":"2023-01-01","paper":"/paper/weakly-supervised-semantic-segmentation-via-3","paper_url":"http://openaccess.thecvf.com//content/CVPR2023/html/Kweon_Weakly_Supervised_Semantic_Segmentation_via_Adversarial_Learning_of_Classifier_and_CVPR_2023_paper.html","paper_title":"Weakly Supervised Semantic Segmentation via Adversarial Learning of Classifier and Reconstructor","code":"https://github.com/sangrockeg/acr","n_code_links":1,"syntology":null},{"rank_in_archive_order":15,"model":"BECO(DeepLabV3Plus+R101)","metrics":{"mIoU":"45.1"},"uses_additional_data":false,"paper_date":"2023-01-01","paper":"/paper/boundary-enhanced-co-training-for-weakly","paper_url":"http://openaccess.thecvf.com//content/CVPR2023/html/Rong_Boundary-Enhanced_Co-Training_for_Weakly_Supervised_Semantic_Segmentation_CVPR_2023_paper.html","paper_title":"Boundary-Enhanced Co-Training for Weakly Supervised Semantic Segmentation","code":"https://github.com/ShenghaiRong/BECO","n_code_links":1,"syntology":null},{"rank_in_archive_order":16,"model":"ViT-PCM","metrics":{"mIoU":"45.0"},"uses_additional_data":false,"paper_date":"2022-10-31","paper":"/paper/max-pooling-with-vision-transformers","paper_url":"https://arxiv.org/abs/2210.17400v1","paper_title":"Max Pooling with Vision Transformers reconciles class and shape in weakly supervised semantic segmentation","code":"https://github.com/deepplants/vit-pcm","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":5,"n_samples":5,"n_pointer_only_licence":0}},{"rank_in_archive_order":17,"model":"ACR-WSSS(DeepLabV2-ResNet101)","metrics":{"mIoU":"45.0"},"uses_additional_data":false,"paper_date":"2023-08-08","paper":"/paper/all-pairs-consistency-learning-for-weakly","paper_url":"https://arxiv.org/abs/2308.04321v2","paper_title":"All-pairs Consistency Learning for Weakly Supervised Semantic Segmentation","code":"https://github.com/opennlplab/acr_wsss","n_code_links":1,"syntology":null},{"rank_in_archive_order":18,"model":"AMN (DeepLabV2-ResNet101)","metrics":{"mIoU":"44.7"},"uses_additional_data":false,"paper_date":"2022-03-30","paper":"/paper/threshold-matters-in-wsss-manipulating-the","paper_url":"https://arxiv.org/abs/2203.16045v1","paper_title":"Threshold Matters in WSSS: Manipulating the Activation for the Robust and Accurate Segmentation Model Against Thresholds","code":"https://github.com/gaviotas/amn","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":1,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":19,"model":"L2G (DeepLabV2-ResNet101)","metrics":{"mIoU":"44.2"},"uses_additional_data":false,"paper_date":"2022-04-07","paper":"/paper/l2g-a-simple-local-to-global-knowledge","paper_url":"https://arxiv.org/abs/2204.03206v1","paper_title":"L2G: A Simple Local-to-Global Knowledge Transfer Framework for Weakly Supervised Semantic Segmentation","code":"https://github.com/pengtaojiang/l2g","n_code_links":1,"syntology":null},{"rank_in_archive_order":20,"model":"RIB (DeepLabV2-ResNet101, No Saliency)","metrics":{"mIoU":"43.8"},"uses_additional_data":false,"paper_date":"2021-10-13","paper":"/paper/reducing-information-bottleneck-for-weakly","paper_url":"https://arxiv.org/abs/2110.06530v1","paper_title":"Reducing Information Bottleneck for Weakly Supervised Semantic Segmentation","code":"https://github.com/jbeomlee93/rib","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":2,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":21,"model":"SIPE(ResNet-38, no saliency, no RW)","metrics":{"mIoU":"43.6"},"uses_additional_data":false,"paper_date":"2022-03-06","paper":"/paper/self-supervised-image-specific-prototype","paper_url":"https://arxiv.org/abs/2203.02909v1","paper_title":"Self-supervised Image-specific Prototype Exploration for Weakly Supervised Semantic Segmentation","code":"https://github.com/chenqi1126/sipe","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":22,"model":"RS+EPM (ResNet-50, single-stage)","metrics":{"mIoU":"42.2"},"uses_additional_data":false,"paper_date":"2022-04-14","paper":"/paper/recurseed-and-certainmix-for-weakly","paper_url":"https://arxiv.org/abs/2204.06754v4","paper_title":"RecurSeed and EdgePredictMix: Pseudo-Label Refinement Learning for Weakly Supervised Semantic Segmentation across Single- and Multi-Stage Frameworks","code":"https://github.com/shjo-april/recurseed_and_edgepredictmix","n_code_links":2,"syntology":null},{"rank_in_archive_order":23,"model":"ClusterCAM","metrics":{"mIoU":"41.8"},"uses_additional_data":false,"paper_date":"2024-01-05","paper":"/paper/clustering-guided-class-activation-for-weakly","paper_url":"https://ieeexplore.ieee.org/abstract/document/10381698","paper_title":"Clustering-Guided Class Activation for Weakly Supervised Semantic Segmentation","code":"https://github.com/DCVL-WSSS/ClusterCAM","n_code_links":1,"syntology":null},{"rank_in_archive_order":24,"model":"URN(Res2Net-101, no saliency, no RW)","metrics":{"mIoU":"41.5"},"uses_additional_data":false,"paper_date":"2021-12-14","paper":"/paper/uncertainty-estimation-via-response-scaling","paper_url":"https://arxiv.org/abs/2112.07431v1","paper_title":"Uncertainty Estimation via Response Scaling for Pseudo-mask Noise Mitigation in Weakly-supervised Semantic Segmentation","code":"https://github.com/xmed-lab/urn","n_code_links":1,"syntology":null},{"rank_in_archive_order":25,"model":"URN(ScaleNet-101, no saliency, no RW)","metrics":{"mIoU":"40.8"},"uses_additional_data":false,"paper_date":"2021-12-14","paper":"/paper/uncertainty-estimation-via-response-scaling","paper_url":"https://arxiv.org/abs/2112.07431v1","paper_title":"Uncertainty Estimation via Response Scaling for Pseudo-mask Noise Mitigation in Weakly-supervised Semantic Segmentation","code":"https://github.com/xmed-lab/urn","n_code_links":1,"syntology":null},{"rank_in_archive_order":26,"model":"URN(ResNet-101, no saliency, no RW)","metrics":{"mIoU":"40.7"},"uses_additional_data":false,"paper_date":"2021-12-14","paper":"/paper/uncertainty-estimation-via-response-scaling","paper_url":"https://arxiv.org/abs/2112.07431v1","paper_title":"Uncertainty Estimation via Response Scaling for Pseudo-mask Noise Mitigation in Weakly-supervised Semantic Segmentation","code":"https://github.com/xmed-lab/urn","n_code_links":1,"syntology":null},{"rank_in_archive_order":27,"model":"URN(ResNet-38, no saliency, no RW)","metrics":{"mIoU":"40.5"},"uses_additional_data":false,"paper_date":"2021-12-14","paper":"/paper/uncertainty-estimation-via-response-scaling","paper_url":"https://arxiv.org/abs/2112.07431v1","paper_title":"Uncertainty Estimation via Response Scaling for Pseudo-mask Noise Mitigation in Weakly-supervised Semantic Segmentation","code":"https://github.com/xmed-lab/urn","n_code_links":1,"syntology":null},{"rank_in_archive_order":28,"model":"PMM(ScaleNet101, no saliency, no RW)","metrics":{"mIoU":"40.2"},"uses_additional_data":false,"paper_date":"2021-08-30","paper":"/paper/pseudo-mask-matters-inweakly-supervised","paper_url":"https://arxiv.org/abs/2108.12995v2","paper_title":"Pseudo-mask Matters in Weakly-supervised Semantic Segmentation","code":"https://github.com/eli-yili/pmm","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":5,"n_samples":5,"n_pointer_only_licence":0}},{"rank_in_archive_order":29,"model":"AFA","metrics":{"mIoU":"38.9"},"uses_additional_data":false,"paper_date":"2022-03-05","paper":"/paper/learning-affinity-from-attention-end-to-end","paper_url":"https://arxiv.org/abs/2203.02664v2","paper_title":"Learning Affinity from Attention: End-to-End Weakly-Supervised Semantic Segmentation with Transformers","code":"https://github.com/rulixiang/afa","n_code_links":2,"syntology":null},{"rank_in_archive_order":30,"model":"PMM(ResNet38, no saliency, no RW)","metrics":{"mIoU":"36.7"},"uses_additional_data":false,"paper_date":"2021-08-30","paper":"/paper/pseudo-mask-matters-inweakly-supervised","paper_url":"https://arxiv.org/abs/2108.12995v2","paper_title":"Pseudo-mask Matters in Weakly-supervised Semantic Segmentation","code":"https://github.com/eli-yili/pmm","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":5,"n_samples":5,"n_pointer_only_licence":0}},{"rank_in_archive_order":31,"model":"OC-CSE(ResNet38, no saliency, no RW)","metrics":{"mIoU":"36.4"},"uses_additional_data":false,"paper_date":"2021-01-01","paper":"/paper/unlocking-the-potential-of-ordinary","paper_url":"http://openaccess.thecvf.com//content/ICCV2021/html/Kweon_Unlocking_the_Potential_of_Ordinary_Classifier_Class-Specific_Adversarial_Erasing_Framework_ICCV_2021_paper.html","paper_title":"Unlocking the Potential of Ordinary Classifier: Class-Specific Adversarial Erasing Framework for Weakly Supervised Semantic Segmentation","code":"https://github.com/kaist-vilab/oc-cse","n_code_links":1,"syntology":null},{"rank_in_archive_order":32,"model":"VWL-L","metrics":{"mIoU":"36.2"},"uses_additional_data":false,"paper_date":"2022-02-10","paper":"/paper/weakly-supervised-semantic-segmentation-with-1","paper_url":"https://arxiv.org/abs/2202.04812v1","paper_title":"Weakly-Supervised Semantic Segmentation with Visual Words Learning and Hybrid Pooling","code":"https://github.com/rulixiang/vwe/tree/master/v2","n_code_links":1,"syntology":null},{"rank_in_archive_order":33,"model":"VWL-M","metrics":{"mIoU":"36.1"},"uses_additional_data":false,"paper_date":"2022-02-10","paper":"/paper/weakly-supervised-semantic-segmentation-with-1","paper_url":"https://arxiv.org/abs/2202.04812v1","paper_title":"Weakly-Supervised Semantic Segmentation with Visual Words Learning and Hybrid Pooling","code":"https://github.com/rulixiang/vwe/tree/master/v2","n_code_links":1,"syntology":null},{"rank_in_archive_order":34,"model":"EPS","metrics":{"mIoU":"35.7"},"uses_additional_data":true,"paper_date":"2021-05-19","paper":"/paper/railroad-is-not-a-train-saliency-as-pseudo-1","paper_url":"https://arxiv.org/abs/2105.08965v1","paper_title":"Railroad is not a Train: Saliency as Pseudo-pixel Supervision for Weakly Supervised Semantic Segmentation","code":"https://github.com/halbielee/EPS","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":35,"model":"SLRNet","metrics":{"mIoU":"35.0"},"uses_additional_data":false,"paper_date":"2022-03-19","paper":"/paper/learning-self-supervised-low-rank-network-for","paper_url":"https://arxiv.org/abs/2203.10278v1","paper_title":"Learning Self-Supervised Low-Rank Network for Single-Stage Weakly and Semi-Supervised Semantic Segmentation","code":"https://github.com/DesertsP/SLRNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":36,"model":"SGAN","metrics":{"mIoU":"33.6"},"uses_additional_data":true,"paper_date":"2019-10-12","paper":"/paper/saliency-guided-self-attention-network-for","paper_url":"https://arxiv.org/abs/1910.05475v2","paper_title":"Saliency Guided Self-attention Network for Weakly and Semi-supervised Semantic Segmentation","code":"https://github.com/yaoqi-zd/SGAN","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":2,"n_samples":5,"n_pointer_only_licence":1}},{"rank_in_archive_order":37,"model":"IRNet+CONTA","metrics":{"mIoU":"33.4"},"uses_additional_data":false,"paper_date":"2020-09-26","paper":"/paper/causal-intervention-for-weakly-supervised","paper_url":"https://arxiv.org/abs/2009.12547v2","paper_title":"Causal Intervention for Weakly-Supervised Semantic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":38,"model":"GroupWSSS","metrics":{"mIoU":"28.4"},"uses_additional_data":true,"paper_date":"2020-12-09","paper":"/paper/group-wise-semantic-mining-for-weakly","paper_url":"https://arxiv.org/abs/2012.05007v1","paper_title":"Group-Wise Semantic Mining for Weakly Supervised Semantic Segmentation","code":"https://github.com/Lixy1997/Group-WSSS","n_code_links":1,"syntology":null},{"rank_in_archive_order":39,"model":"DSRG","metrics":{"mIoU":"26.0"},"uses_additional_data":true,"paper_date":"2018-06-01","paper":"/paper/weakly-supervised-semantic-segmentation","paper_url":"http://openaccess.thecvf.com/content_cvpr_2018/html/Huang_Weakly-Supervised_Semantic_Segmentation_CVPR_2018_paper.html","paper_title":"Weakly-Supervised Semantic Segmentation Network With Deep Seeded Region Growing","code":"https://github.com/speedinghzl/DSRG","n_code_links":1,"syntology":null}],"since_archive":{"present":false,"note":"No Syntology-extracted rows are published in this build."},"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":15,"rows_with_any_sample_ran":9,"distinct_papers_with_graph_line":13,"distinct_papers_with_any_sample_ran":8,"samples_over_distinct_papers":{"n_ran":24,"n_unverified":25,"n_samples":49,"n_pointer_only_licence":19,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":25,"n_unverified":30,"n_samples":55,"n_pointer_only_licence":19,"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"}}}