TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

SotA/Computer Vision/Panoptic Segmentation/Cityscapes val

Panoptic Segmentation on Cityscapes val

Metric: AP (higher is better)

LeaderboardDataset
Loading chart...

Results

Submit a result
#Model↕AP▼Extra DataPaperDate↕Code
1ViT-P (OneFormer, InternImage-H)50.6NoThe Missing Point in Vision Transformers for Uni...2025-05-26Code
2OneFormer (ConvNeXt-L, single-scale, 512x1024, Mapillary Vistas-pretrained)48.7YesOneFormer: One Transformer to Rule Universal Ima...2022-11-10Code
3Panoptic-DeepLab (SWideRNet [1, 1, 4.5], Mapillary Vistas, multi-scale)46.8YesScaling Wide Residual Networks for Panoptic Segm...2020-11-23-
4OneFormer (ConvNeXt-XL, single-scale)46.7NoOneFormer: One Transformer to Rule Universal Ima...2022-11-10Code
5OneFormer (ConvNeXt-L, single-scale)46.5NoOneFormer: One Transformer to Rule Universal Ima...2022-11-10Code
6AFF-Base (single-scale, point-based Mask2Former)46.2NoAutoFocusFormer: Image Segmentation off the Grid2023-04-24Code
7OneFormer (DiNAT-L, single-scale)45.6NoOneFormer: One Transformer to Rule Universal Ima...2022-11-10Code
8OneFormer (Swin-L, single-scale)45.6NoOneFormer: One Transformer to Rule Universal Ima...2022-11-10Code
9DiNAT-L (Mask2Former)44.5NoDilated Neighborhood Attention Transformer2022-09-29Code
10Axial-DeepLab-XL (Mapillary Vistas, multi-scale)44.2YesAxial-DeepLab: Stand-Alone Axial-Attention for P...2020-03-17Code
11AFF-Small (single-scale, point-based Mask2Former)44.2NoAutoFocusFormer: Image Segmentation off the Grid2023-04-24Code
12kMaX-DeepLab (single-scale)44NokMaX-DeepLab: k-means Mask Transformer2022-07-08Code
13Mask2Former (Swin-L)43.6NoMasked-attention Mask Transformer for Universal ...2021-12-02Code
14EfficientPS43.5YesEfficientPS: Efficient Panoptic Segmentation2020-04-05Code
15Panoptic-DeepLab (SWideRNet [1, 1, 4.5], Mapillary Vistas, single-scale)42.8YesScaling Wide Residual Networks for Panoptic Segm...2020-11-23-
16EfficientPS (Cityscapes-fine)39.1NoEfficientPS: Efficient Panoptic Segmentation2020-04-05Code
17UPSNet (ResNet-101, multiscale)39YesUPSNet: A Unified Panoptic Segmentation Network2019-01-12Code
18TASCNet (ResNet-50, multi-scale)39YesLearning to Fuse Things and Stuff2018-12-04-
19Panoptic-DeepLab (X71)38.5YesPanoptic-DeepLab: A Simple, Strong, and Fast Bas...2019-11-22Code
20UPSNet (ResNet-101)37.8YesUPSNet: A Unified Panoptic Segmentation Network2019-01-12Code
21TASCNet (ResNet-50)37.6YesLearning to Fuse Things and Stuff2018-12-04-
22MRCNN + PSPNet (ResNet-101)36.4YesPanoptic Segmentation2018-01-03Code
23AdaptIS (ResNeXt-101)36.3NoAdaptIS: Adaptive Instance Selection Network2019-09-17-
24AUNet (ResNet-101-FPN)34.4NoAttention-guided Unified Network for Panoptic Se...2018-12-10-
25COPS (ResNet-50)34.1NoCombinatorial Optimization for Panoptic Segmenta...2021-06-06Code
26AdaptIS (ResNet-101)33.9NoAdaptIS: Adaptive Instance Selection Network2019-09-17-
27UPSNet (ResNet-50)33.3NoUPSNet: A Unified Panoptic Segmentation Network2019-01-12Code
28Panoptic FPN (ResNet-101)33NoPanoptic Feature Pyramid Networks2019-01-08Code
29AdaptIS (ResNet-50)32.3NoAdaptIS: Adaptive Instance Selection Network2019-09-17-
30Dynamically Instantiated Network (ResNet-101)28.6NoWeakly- and Semi-Supervised Panoptic Segmentation2018-08-10Code