{"url":"/sota/panoptic-segmentation-on-coco-test-dev","task":{"name":"Panoptic Segmentation","url":"/task/panoptic-segmentation","note":null},"dataset":{"name":"COCO test-dev","url":"/dataset/coco"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"**Panoptic Segmentation** is a computer vision task that combines semantic segmentation and instance segmentation to provide a comprehensive understanding of the scene. The goal of panoptic segmentation is to segment the image into semantically meaningful parts or regions, while also detecting and distinguishing individual instances of objects within those regions. In a given image, every pixel is assigned a semantic label, and pixels belonging to \"things\" classes (countable objects with instances, like cars and people) are assigned unique instance IDs.\r\n<span style=\"color:grey; opacity: 0.6\">( Image credit: [Detectron2](https://github.com/facebookresearch/detectron2) )</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":["PQ","PQst","PQth"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"PQ":null,"PQst":null,"PQth":null}},"counts":{"rows":38,"rows_with_code":27,"rows_with_paper_page":38,"rows_dated":38,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"Mask DINO (single scale)","metrics":{"PQ":"59.5","PQst":"-","PQth":"-"},"uses_additional_data":false,"paper_date":"2022-06-06","paper":"/paper/mask-dino-towards-a-unified-transformer-based-1","paper_url":"https://arxiv.org/abs/2206.02777v3","paper_title":"Mask DINO: Towards A Unified Transformer-based Framework for Object Detection and Segmentation","code":"https://github.com/PaddlePaddle/PaddleDetection","n_code_links":10,"syntology":{"n_ran":11,"n_unverified":2,"n_samples":13,"n_pointer_only_licence":13}},{"rank_in_archive_order":2,"model":"kMaX-DeepLab (single-scale)","metrics":{"PQ":"58.5","PQst":"49.0","PQth":"64.8"},"uses_additional_data":false,"paper_date":"2022-07-08","paper":"/paper/k-means-mask-transformer","paper_url":"https://arxiv.org/abs/2207.04044v5","paper_title":"kMaX-DeepLab: k-means Mask Transformer","code":"https://github.com/google-research/deeplab2","n_code_links":3,"syntology":null},{"rank_in_archive_order":3,"model":"Mask2Former (Swin-L)","metrics":{"PQ":"58.3","PQst":"48.1","PQth":"65.1"},"uses_additional_data":false,"paper_date":"2021-12-02","paper":"/paper/masked-attention-mask-transformer-for","paper_url":"https://arxiv.org/abs/2112.01527v3","paper_title":"Masked-attention Mask Transformer for Universal Image Segmentation","code":"https://github.com/huggingface/transformers","n_code_links":7,"syntology":{"n_ran":2,"n_unverified":6,"n_samples":8,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"Panoptic SegFormer (Swin-L)","metrics":{"PQ":"56.2","PQst":"47.0","PQth":"62.3"},"uses_additional_data":false,"paper_date":"2021-09-08","paper":"/paper/panoptic-segformer","paper_url":"https://arxiv.org/abs/2109.03814v4","paper_title":"Panoptic SegFormer: Delving Deeper into Panoptic Segmentation with Transformers","code":"https://github.com/zhiqi-li/Panoptic-SegFormer","n_code_links":3,"syntology":null},{"rank_in_archive_order":5,"model":"Panoptic SegFormer (PVTv2-B5)","metrics":{"PQ":"55.8","PQst":"46.5","PQth":"61.9"},"uses_additional_data":false,"paper_date":"2021-09-08","paper":"/paper/panoptic-segformer","paper_url":"https://arxiv.org/abs/2109.03814v4","paper_title":"Panoptic SegFormer: Delving Deeper into Panoptic Segmentation with Transformers","code":"https://github.com/zhiqi-li/Panoptic-SegFormer","n_code_links":3,"syntology":null},{"rank_in_archive_order":6,"model":"CMT-DeepLab (single-scale)","metrics":{"PQ":"55.7","PQst":"46.8","PQth":"61.6"},"uses_additional_data":false,"paper_date":"2022-06-17","paper":"/paper/cmt-deeplab-clustering-mask-transformers-for-1","paper_url":"https://arxiv.org/abs/2206.08948v1","paper_title":"CMT-DeepLab: Clustering Mask Transformers for Panoptic Segmentation","code":"https://github.com/bytedance/kmax-deeplab","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":8,"n_samples":8,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"K-Net (Swin-L)","metrics":{"PQ":"55.2","PQst":"46.2","PQth":"61.2"},"uses_additional_data":false,"paper_date":"2021-06-28","paper":"/paper/k-net-towards-unified-image-segmentation","paper_url":"https://arxiv.org/abs/2106.14855v2","paper_title":"K-Net: Towards Unified Image Segmentation","code":"https://github.com/zwwwayne/k-net","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":8,"model":"MaskConver (ResNet50, single-scale)","metrics":{"PQ":"53.6","PQst":"58.9","PQth":"45.6"},"uses_additional_data":false,"paper_date":"2023-12-11","paper":"/paper/maskconver-revisiting-pure-convolution-model","paper_url":"https://arxiv.org/abs/2312.06052v1","paper_title":"MaskConver: Revisiting Pure Convolution Model for Panoptic Segmentation","code":"https://github.com/tensorflow/models","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"MaskFormer (Swin-L)","metrics":{"PQ":"53.3","PQst":"44.5","PQth":"59.1"},"uses_additional_data":false,"paper_date":"2021-07-13","paper":"/paper/per-pixel-classification-is-not-all-you-need","paper_url":"https://arxiv.org/abs/2107.06278v2","paper_title":"Per-Pixel Classification is Not All You Need for Semantic Segmentation","code":"https://github.com/huggingface/transformers","n_code_links":3,"syntology":null},{"rank_in_archive_order":10,"model":"Panoptic FCN* (Swin-L)","metrics":{"PQ":"52.7","PQth":" 59.4"},"uses_additional_data":false,"paper_date":"2020-12-01","paper":"/paper/fully-convolutional-networks-for-panoptic","paper_url":"https://arxiv.org/abs/2012.00720v2","paper_title":"Fully Convolutional Networks for Panoptic Segmentation","code":"https://github.com/dvlab-research/panopticfcn","n_code_links":6,"syntology":null},{"rank_in_archive_order":11,"model":"REFINE (ResNeXt-101-DCN)","metrics":{"PQ":"51.5","PQst":"39.2","PQth":"59.6"},"uses_additional_data":false,"paper_date":"2020-12-15","paper":"/paper/refine-prediction-fusion-network-for-panoptic","paper_url":"https://sites.google.com/view/refine-aaai2021/home","paper_title":"REFINE: Prediction Fusion Network for Panoptic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":12,"model":"MaX-DeepLab-L (single-scale)","metrics":{"PQ":"51.3","PQst":"42.4","PQth":"57.2"},"uses_additional_data":false,"paper_date":"2020-12-01","paper":"/paper/max-deeplab-end-to-end-panoptic-segmentation","paper_url":"https://arxiv.org/abs/2012.00759v3","paper_title":"MaX-DeepLab: End-to-End Panoptic Segmentation with Mask Transformers","code":"https://github.com/google-research/deeplab2","n_code_links":3,"syntology":null},{"rank_in_archive_order":13,"model":"Panoptic SegFormer (ResNet-101)","metrics":{"PQ":"50.9","PQst":"43.0","PQth":"56.2"},"uses_additional_data":false,"paper_date":"2021-09-08","paper":"/paper/panoptic-segformer","paper_url":"https://arxiv.org/abs/2109.03814v4","paper_title":"Panoptic SegFormer: Delving Deeper into Panoptic Segmentation with Transformers","code":"https://github.com/zhiqi-li/Panoptic-SegFormer","n_code_links":3,"syntology":null},{"rank_in_archive_order":14,"model":"Panoptic SegFormer (ResNet-50)","metrics":{"PQ":"50.2","PQst":"42.4","PQth":"55.3"},"uses_additional_data":false,"paper_date":"2021-09-08","paper":"/paper/panoptic-segformer","paper_url":"https://arxiv.org/abs/2109.03814v4","paper_title":"Panoptic SegFormer: Delving Deeper into Panoptic Segmentation with Transformers","code":"https://github.com/zhiqi-li/Panoptic-SegFormer","n_code_links":3,"syntology":null},{"rank_in_archive_order":15,"model":"DetectoRS (ResNeXt-101-64x4d, multi-scale)","metrics":{"PQ":"50","PQst":"37.2","PQth":"58.5"},"uses_additional_data":false,"paper_date":"2020-06-03","paper":"/paper/detectors-detecting-objects-with-recursive-1","paper_url":"https://arxiv.org/abs/2006.02334v2","paper_title":"DetectoRS: Detecting Objects with Recursive Feature Pyramid and Switchable Atrous Convolution","code":"https://github.com/open-mmlab/mmdetection","n_code_links":6,"syntology":{"n_ran":0,"n_unverified":4,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":16,"model":"REFINE (ResNet-101-DCN)","metrics":{"PQ":"49.6","PQst":"37.7","PQth":"57.5"},"uses_additional_data":false,"paper_date":"2020-12-15","paper":"/paper/refine-prediction-fusion-network-for-panoptic","paper_url":"https://sites.google.com/view/refine-aaai2021/home","paper_title":"REFINE: Prediction Fusion Network for Panoptic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":17,"model":"SpatialFlow(ResNet-101-FPN)","metrics":{"PQ":"48.5","PQst":"37.9","PQth":"55.5"},"uses_additional_data":false,"paper_date":"2019-10-19","paper":"/paper/spatialflow-bridging-all-tasks-for-panoptic","paper_url":"https://arxiv.org/abs/1910.08787v3","paper_title":"SpatialFlow: Bridging All Tasks for Panoptic Segmentation","code":"https://github.com/chensnathan/SpatialFlow","n_code_links":1,"syntology":null},{"rank_in_archive_order":18,"model":"Ada-Segment (ResNet-101-DCN)","metrics":{"PQ":"48.5","PQst":"37.6","PQth":"55.7"},"uses_additional_data":false,"paper_date":"2020-12-07","paper":"/paper/ada-segment-automated-multi-loss-adaptation","paper_url":"https://arxiv.org/abs/2012.03603v1","paper_title":"Ada-Segment: Automated Multi-loss Adaptation for Panoptic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":19,"model":"K-Net (R101-FPN-DCN)","metrics":{"PQ":"48.3","PQst":"39.7","PQth":"54"},"uses_additional_data":false,"paper_date":"2021-06-28","paper":"/paper/k-net-towards-unified-image-segmentation","paper_url":"https://arxiv.org/abs/2106.14855v2","paper_title":"K-Net: Towards Unified Image Segmentation","code":"https://github.com/zwwwayne/k-net","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":20,"model":"SOGNet (ResNet-101-FPN)","metrics":{"PQ":"47.8"},"uses_additional_data":false,"paper_date":"2019-11-18","paper":"/paper/sognet-scene-overlap-graph-network-for","paper_url":"https://arxiv.org/abs/1911.07527v1","paper_title":"SOGNet: Scene Overlap Graph Network for Panoptic Segmentation","code":"https://github.com/LaoYang1994/SOGNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":21,"model":"Panoptic FCN*++ (DCN-101-FPN)","metrics":{"PQ":"47.5","PQst":"38.2","PQth":"53.7"},"uses_additional_data":false,"paper_date":"2020-12-01","paper":"/paper/fully-convolutional-networks-for-panoptic","paper_url":"https://arxiv.org/abs/2012.00720v2","paper_title":"Fully Convolutional Networks for Panoptic Segmentation","code":"https://github.com/dvlab-research/panopticfcn","n_code_links":6,"syntology":null},{"rank_in_archive_order":22,"model":"UPSNet (ResNet-101-FPN)","metrics":{"PQ":"46.6","PQst":"36.7","PQth":"53.2"},"uses_additional_data":false,"paper_date":"2019-01-12","paper":"/paper/upsnet-a-unified-panoptic-segmentation","paper_url":"http://arxiv.org/abs/1901.03784v2","paper_title":"UPSNet: A Unified Panoptic Segmentation Network","code":"https://github.com/uber-research/UPSNet","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":23,"model":"OCFusion (ResNeXt-101-FPN)","metrics":{"PQ":"46.6","PQst":"35.7","PQth":"54.0"},"uses_additional_data":false,"paper_date":"2019-06-13","paper":"/paper/learning-instance-occlusion-for-panoptic","paper_url":"https://arxiv.org/abs/1906.05896v4","paper_title":"Learning Instance Occlusion for Panoptic Segmentation","code":"https://github.com/jlazarow/learning_instance_occlusion","n_code_links":1,"syntology":null},{"rank_in_archive_order":24,"model":"Panoptic-DeepLab (SWideRNet-[1, 1, 4], multi-scale)","metrics":{"PQ":"46.5","PQst":"38.2","PQth":"52.0"},"uses_additional_data":false,"paper_date":"2020-11-23","paper":"/paper/scaling-wide-residual-networks-for-panoptic","paper_url":"https://arxiv.org/abs/2011.11675v2","paper_title":"Scaling Wide Residual Networks for Panoptic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":25,"model":"AUNet (ResNext-152-FPN)","metrics":{"PQ":"46.5","PQst":"32.5","PQth":"55.8"},"uses_additional_data":false,"paper_date":"2018-12-10","paper":"/paper/attention-guided-unified-network-for-panoptic","paper_url":"http://arxiv.org/abs/1812.03904v2","paper_title":"Attention-guided Unified Network for Panoptic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":26,"model":"AUNet (ResNet-152-FPN)","metrics":{"PQ":"45.5","PQst":"31.6","PQth":"54.7"},"uses_additional_data":false,"paper_date":"2018-12-10","paper":"/paper/attention-guided-unified-network-for-panoptic","paper_url":"http://arxiv.org/abs/1812.03904v2","paper_title":"Attention-guided Unified Network for Panoptic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":27,"model":"AUNet (ResNet-101-FPN)","metrics":{"PQ":"45.2","PQst":"31.3","PQth":"54.4"},"uses_additional_data":false,"paper_date":"2018-12-10","paper":"/paper/attention-guided-unified-network-for-panoptic","paper_url":"http://arxiv.org/abs/1812.03904v2","paper_title":"Attention-guided Unified Network for Panoptic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":28,"model":"Axial-DeepLab-L (multi-scale)","metrics":{"PQ":"44.2","PQst":"36.8","PQth":"49.2"},"uses_additional_data":false,"paper_date":"2020-03-17","paper":"/paper/axial-deeplab-stand-alone-axial-attention-for","paper_url":"https://arxiv.org/abs/2003.07853v2","paper_title":"Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation","code":"https://github.com/The-AI-Summer/self_attention","n_code_links":5,"syntology":{"n_ran":3,"n_unverified":8,"n_samples":11,"n_pointer_only_licence":1}},{"rank_in_archive_order":29,"model":"Axial-DeepLab-L","metrics":{"PQ":"43.6","PQst":"35.6","PQth":"48.9"},"uses_additional_data":false,"paper_date":"2020-03-17","paper":"/paper/axial-deeplab-stand-alone-axial-attention-for","paper_url":"https://arxiv.org/abs/2003.07853v2","paper_title":"Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation","code":"https://github.com/The-AI-Summer/self_attention","n_code_links":5,"syntology":{"n_ran":3,"n_unverified":8,"n_samples":11,"n_pointer_only_licence":1}},{"rank_in_archive_order":30,"model":"AdaptIS (ResNeXt-101)","metrics":{"PQ":"42.8","PQst":"31.8","PQth":"50.1"},"uses_additional_data":false,"paper_date":"2019-09-17","paper":"/paper/adaptis-adaptive-instance-selection-network","paper_url":"https://arxiv.org/abs/1909.07829v1","paper_title":"AdaptIS: Adaptive Instance Selection Network","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":31,"model":"Panoptic-DeepLab (Xception-71)","metrics":{"PQ":"41.4","PQst":"35.9","PQth":"45.1"},"uses_additional_data":false,"paper_date":"2019-11-22","paper":"/paper/panoptic-deeplab-a-simple-strong-and-fast","paper_url":"https://arxiv.org/abs/1911.10194v3","paper_title":"Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up Panoptic Segmentation","code":"https://github.com/tensorflow/models/tree/master/official/projects/panoptic","n_code_links":9,"syntology":{"n_ran":3,"n_unverified":6,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":32,"model":"Panoptic FPN","metrics":{"PQ":"40.9","PQst":"29.7","PQth":"48.3"},"uses_additional_data":false,"paper_date":"2019-01-08","paper":"/paper/panoptic-feature-pyramid-networks","paper_url":"http://arxiv.org/abs/1901.02446v2","paper_title":"Panoptic Feature Pyramid Networks","code":"https://github.com/facebookresearch/detectron2","n_code_links":12,"syntology":{"n_ran":7,"n_unverified":5,"n_samples":12,"n_pointer_only_licence":0}},{"rank_in_archive_order":33,"model":"TASCNet","metrics":{"PQ":"40.7","PQst":"31.0","PQth":"47.0"},"uses_additional_data":false,"paper_date":"2018-12-04","paper":"/paper/learning-to-fuse-things-and-stuff","paper_url":"https://arxiv.org/abs/1812.01192v2","paper_title":"Learning to Fuse Things and Stuff","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":34,"model":"EPSNet (ResNet-101-FPN)","metrics":{"PQ":"38.9","PQst":"31.0","PQth":"44.1"},"uses_additional_data":false,"paper_date":"2020-03-23","paper":"/paper/epsnet-efficient-panoptic-segmentation","paper_url":"https://arxiv.org/abs/2003.10142v3","paper_title":"EPSNet: Efficient Panoptic Segmentation Network with Cross-layer Attention Fusion","code":"https://github.com/neo85824/epsnet","n_code_links":1,"syntology":null},{"rank_in_archive_order":35,"model":"COPS (ResNet-50)","metrics":{"PQ":"38.5","PQst":"34.8","PQth":"41.0"},"uses_additional_data":false,"paper_date":"2021-06-06","paper":"/paper/combinatorial-optimization-for-panoptic","paper_url":"https://arxiv.org/abs/2106.03188v3","paper_title":"Combinatorial Optimization for Panoptic Segmentation: A Fully Differentiable Approach","code":"https://github.com/LPMP/LPMP","n_code_links":2,"syntology":null},{"rank_in_archive_order":36,"model":"PCV (ResNet-50)","metrics":{"PQ":"37.7","PQst":"33.1","PQth":"40.7"},"uses_additional_data":false,"paper_date":"2020-04-04","paper":"/paper/pixel-consensus-voting-for-panoptic","paper_url":"https://arxiv.org/abs/2004.01849v1","paper_title":"Pixel Consensus Voting for Panoptic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":37,"model":"GES Net","metrics":{"PQ":"33.7","PQst":"31.5","PQth":"35.1"},"uses_additional_data":false,"paper_date":"2019-08-24","paper":"/paper/generator-evaluator-selector-net-a-modular","paper_url":"https://arxiv.org/abs/1908.09108v4","paper_title":"Generator evaluator-selector net for panoptic image segmentation and splitting unfamiliar objects into parts","code":"https://github.com/sagieppel/Generator-evaluator-selector-net-a-modular-approach-for-panoptic-segmentation","n_code_links":2,"syntology":null},{"rank_in_archive_order":38,"model":"JSIS-Net","metrics":{"PQ":"27.2","PQst":"23.4","PQth":"29.6"},"uses_additional_data":false,"paper_date":"2018-09-06","paper":"/paper/panoptic-segmentation-with-a-joint-semantic","paper_url":"http://arxiv.org/abs/1809.02110v2","paper_title":"Panoptic Segmentation with a Joint Semantic and Instance Segmentation Network","code":null,"n_code_links":0,"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":11,"rows_with_any_sample_ran":7,"distinct_papers_with_graph_line":9,"distinct_papers_with_any_sample_ran":6,"samples_over_distinct_papers":{"n_ran":29,"n_unverified":40,"n_samples":69,"n_pointer_only_licence":17,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":32,"n_unverified":49,"n_samples":81,"n_pointer_only_licence":18,"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"}}}