Semantic Segmentation on PASCAL VOC 2012 (test)
94.1mIoUZegCLIP
Evaluation results
| Method | Links | |||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ZegCLIPsource_dataset=COCO, self-training=true2022.12 | 94.1 | 96.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 97.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ZegCLIPsource_dataset=COCO2022.12 | 93.6 | 96.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 96.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SegNeXt-LBackbone=MSCAN-L, Pre-training=COCO2022.09 | 90.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Eff-L2 w/ Self-trainingPre-trained=ImageNet++, FLOPs=229B, Params=485M, Self-training=true, fine tuning=train+val2020.06 | 90.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| NAS-FPNBackbone=EfficientNet-L2, Pre-training=JFT-300M and 300M unlabeled images2022.09 | 90.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLabv3+Pre-trained=ImageNet, JFT, COCO, FLOPs=3055B, multi-scale/flip ensembling=true, fine tuning=train+val2020.06 | 89 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLabV3+Backbone=Xception-JFT, Pre-training=JFT-300M and 300M unlabeled images2022.09 | 89 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RecoNetBackbone=ResNet-101, COCO pre-training=true2026.06 | 88.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MRFM + EALPre-trained=MS-COCO, Backbone=Xception652020.11 | 88.4 | — | — | 78.6 | 97.1 | 80.6 | 89.7 | 97.3 | 93.6 | 96.7 | 59 | 95.4 | 81.1 | 93.2 | 97.5 | 94.2 | 92.9 | 72.3 | 93.1 | 74.2 | 91 | 85 | — | 97.1 | — | — | — | — | — | — | — | — | — | — | — | |
| Ourspre-training=MS-COCO2020.11 | 88.4 | — | — | 78.6 | 97.1 | 80.6 | 89.7 | 97.3 | 93.6 | 96.7 | 59 | 95.4 | 81.1 | 93.2 | 97.5 | 94.2 | 92.9 | 72.3 | 93.1 | 74.2 | 91 | 85 | — | 97.1 | — | — | — | — | — | — | — | — | — | — | — | |
| MSCIImageNet Pretrained=true, COCO Pretrained=true2019.01 | 88 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MSCIPre-trained=MS-COCO2020.11 | 88 | — | — | 76.8 | 97 | 80.6 | 89.3 | 97.4 | 93.8 | 97.1 | 56.7 | 94.3 | 78.3 | 93.5 | 97.1 | 94 | 92.8 | 72.3 | 92.6 | 73.6 | 90.8 | 85.4 | — | 96.8 | — | — | — | — | — | — | — | — | — | — | — | |
| MSCIpre-training=MS-COCO2020.11 | 88 | — | — | 76.8 | 97 | 80.6 | 89.3 | 97.4 | 93.8 | 97.1 | 56.7 | 94.3 | 78.3 | 93.5 | 97.1 | 94 | 92.8 | 72.3 | 92.6 | 73.6 | 90.8 | 85.4 | — | 96.8 | — | — | — | — | — | — | — | — | — | — | — | |
| ExFusebackbone=ResNeXt-131, post-processing=None2018.04 | 87.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ExFusePre-trained=ImageNet, COCO, multi-scale/flip ensembling=true, fine tuning=train+val2020.06 | 87.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ExFusePre-trained=MS-COCO2020.11 | 87.9 | — | — | 80.3 | 97 | 82.5 | 87.8 | 96.3 | 92.6 | 96.4 | 53.3 | 94.3 | 78.4 | 94.1 | 94.9 | 91.6 | 92.3 | 81.7 | 94.8 | 70.3 | 90.1 | 83.8 | — | 96.8 | — | — | — | — | — | — | — | — | — | — | — | |
| ExFusepre-training=MS-COCO2020.11 | 87.9 | — | — | 80.3 | 97 | 82.5 | 87.8 | 96.3 | 92.6 | 96.4 | 53.3 | 94.3 | 78.4 | 94.1 | 94.9 | 91.6 | 92.3 | 81.7 | 94.8 | 70.3 | 90.1 | 83.8 | — | 96.8 | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLabv3+backbone=Xception2018.04 | 87.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLabv3+ImageNet Pretrained=true, COCO Pretrained=true2019.01 | 87.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLabV3+Feature Extractor=Xception-65, Fully Supervised Model (Backbone)=Xception-65, Supervision=F2020.11 | 87.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Deeplabv3+Pre-trained=MS-COCO2020.11 | 87.8 | — | — | 77.1 | 97.1 | 79.3 | 89.3 | 97.4 | 93.2 | 96.6 | 56.9 | 95 | 79.2 | 93.1 | 97 | 94 | 92.8 | 71.3 | 92.9 | 72.4 | 91 | 84.9 | — | 97 | — | — | — | — | — | — | — | — | — | — | — | |
| Deeplabv3+pre-training=MS-COCO2020.11 | 87.8 | — | — | 77.1 | 97.1 | 79.3 | 89.3 | 97.4 | 93.2 | 96.6 | 56.9 | 95 | 79.2 | 93.1 | 97 | 94 | 92.8 | 71.3 | 92.9 | 72.4 | 91 | 84.9 | — | 97 | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLabV3+Backbone=Xception-71, Pre-training=COCO2022.09 | 87.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLabv3+Backbone=Xception-65, COCO pre-training=true2026.06 | 87.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| EMANetBackbone=ResNet101, Pre-training=COCO2022.09 | 87.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| EMANetBackbone=ResNet-101, COCO pre-training=true2026.06 | 87.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SegNeXt-BBackbone=MSCAN-B2022.09 | 87.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CFNetBackbone=ResNet-101, COCO pre-training=true2026.06 | 87.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLabv3-JFTPre-training=JFT-300M2017.06 | 86.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DIS2017.06 | 86.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SDN+Training Data=Using VOC+COCO data2017.08 | 86.6 | — | — | 78.6 | 96 | 79.6 | 84.1 | 97.1 | 91.9 | 96.6 | 48.5 | 94.3 | 78.9 | 93.6 | 95.5 | 92.1 | 91.1 | 75 | 93.8 | 64.8 | 89 | 84.6 | — | 96.9 | — | — | — | — | — | — | — | — | — | — | — | |
| CASIA_IVA_SDN2017.06 | 86.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SDN2018.04 | 86.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SDNPre-trained=MS-COCO2020.11 | 86.6 | — | — | 78.6 | 96 | 79.6 | 84.1 | 97.1 | 91.9 | 96.6 | 48.5 | 94.3 | 78.9 | 93.6 | 95.5 | 92.1 | 91.1 | 75 | 93.8 | 64.8 | 89 | 84.6 | — | 96.9 | — | — | — | — | — | — | — | — | — | — | — | |
| SDNpre-training=MS-COCO2020.11 | 86.6 | — | — | 78.6 | 96 | 79.6 | 84.1 | 97.1 | 91.9 | 96.6 | 48.5 | 94.3 | 78.9 | 93.6 | 95.5 | 92.1 | 91.1 | 75 | 93.8 | 64.8 | 89 | 84.6 | — | 96.9 | — | — | — | — | — | — | — | — | — | — | — | |
| ModuSegTrain status=Training-Free, Method category=Multi-stage2026.04 | 86.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| IDW-CNN2017.06 | 86.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DFNPre-trained on MS-COCO=true2018.04 | 86.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ExFusebackbone=ResNet-101, post-processing=None2018.04 | 86.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DFNPre-trained=MS-COCO2020.11 | 86.2 | — | — | 78.6 | 95.5 | 79.1 | 86.4 | 97.1 | 91.4 | 95 | 47.7 | 92.9 | 77.2 | 91 | 96.7 | 92.2 | 91.7 | 76.5 | 93.1 | 64.4 | 88.3 | 81.2 | — | 96.4 | — | — | — | — | — | — | — | — | — | — | — | |
| DFNpre-training=MS-COCO2020.11 | 86.2 | — | — | 78.6 | 95.5 | 79.1 | 86.4 | 97.1 | 91.4 | 95 | 47.7 | 92.9 | 77.2 | 91 | 96.7 | 92.2 | 91.7 | 76.5 | 93.1 | 64.4 | 88.3 | 81.2 | — | 96.4 | — | — | — | — | — | — | — | — | — | — | — | |
| DFNBackbone=ResNet-101, COCO pre-training=true2026.06 | 86.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| EncNetPre-training=COCO, FPS=12, #Param=54.5M2018.11 | 85.9 | — | — | 76.9 | 94.2 | 80.2 | 85.2 | 96.5 | 90.8 | 96.3 | 47.9 | 93.9 | 77.8 | 88.2 | 95.5 | 90 | 79.4 | 90.2 | 94.2 | 92.8 | 91 | 92.1 | — | 95.3 | — | — | — | — | — | — | — | — | — | — | — | |
| EncNetPre-trained=MS-COCO2020.11 | 85.9 | — | — | 76.9 | 94.2 | 80.2 | 85.3 | 96.5 | 90.8 | 96.3 | 47.9 | 93.9 | 80 | 92.4 | 96.6 | 90.5 | 91.5 | 70.9 | 93.6 | 66.5 | 87.7 | 80.8 | — | 95.3 | — | — | — | — | — | — | — | — | — | — | — | |
| EncNetpre-training=MS-COCO2020.11 | 85.9 | — | — | 76.9 | 94.2 | 80.2 | 85.3 | 96.5 | 90.8 | 96.3 | 47.9 | 93.9 | 80 | 92.4 | 96.6 | 90.5 | 91.5 | 70.9 | 93.6 | 66.5 | 87.7 | 80.8 | — | 95.3 | — | — | — | — | — | — | — | — | — | — | — | |
| HamNetBackbone=ResNet1012022.09 | 85.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| EncNetBackbone=ResNet101, Pre-training=COCO2022.09 | 85.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| EncNetBackbone=ResNet-101, COCO pre-training=true2026.06 | 85.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLabv3Training Data=Using VOC+COCO data2017.08 | 85.7 | — | — | 76.6 | 92.7 | 77.8 | 87.6 | 96.7 | 90.2 | 95.4 | 47.5 | 93.4 | 76.3 | 91.4 | 97.2 | 91 | 92.1 | 71.3 | 90.9 | 68.9 | 90.8 | 79.3 | — | 96.4 | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLabv3Pre-training=MS-COCO2017.06 | 85.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Deeplab v3Pre-trained on MS-COCO=true2018.04 | 85.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLabv32018.04 | 85.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLabv3Pre-training=COCO, FPS=8, #Param=58.0M2018.11 | 85.7 | — | — | 76.6 | 92.7 | 77.8 | 87.6 | 96.7 | 90.2 | 95.4 | 47.5 | 93.4 | 76.3 | 91.4 | 97.2 | 93.9 | 80 | 92.4 | 96.6 | 89.5 | 68.2 | 83.6 | — | 96.4 | — | — | — | — | — | — | — | — | — | — | — | |
| Deeplabv3Pre-trained=MS-COCO2020.11 | 85.7 | — | — | 76.6 | 92.7 | 77.8 | 87.6 | 96.7 | 90.2 | 95.4 | 47.5 | 93.4 | 76.3 | 91.4 | 97.2 | 91 | 92.1 | 71.3 | 90.9 | 68.9 | 90.8 | 79.3 | — | 96.4 | — | — | — | — | — | — | — | — | — | — | — | |
| Deeplabv3pre-training=MS-COCO2020.11 | 85.7 | — | — | 76.6 | 92.7 | 77.8 | 87.6 | 96.7 | 90.2 | 95.4 | 47.5 | 93.4 | 76.3 | 91.4 | 97.2 | 91 | 92.1 | 71.3 | 90.9 | 68.9 | 90.8 | 79.3 | — | 96.4 | — | — | — | — | — | — | — | — | — | — | — | |
| Auto-DeepLab-LImageNet Pretrained=false, COCO Pretrained=true2019.01 | 85.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Zegformersource_dataset=COCO2022.12 | 85.6 | 92.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 92.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PSPNet#Params=65.72018.03 | 85.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PSPNetTraining Data=Using VOC+COCO data2017.08 | 85.4 | — | — | 72.7 | 95 | 78.9 | 84.4 | 94.7 | 92 | 95.7 | 43.1 | 91 | 80.3 | 91.3 | 96.3 | 92.3 | 90.1 | 71.5 | 94.4 | 66.9 | 88.8 | 82 | — | 95.8 | — | — | — | — | — | — | — | — | — | — | — | |
| PSPNet2017.06 | 85.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PSPNetPre-trained on MS-COCO=true2018.04 | 85.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PSPNet2018.04 | 85.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PSPNetImageNet Pretrained=true, COCO Pretrained=true2019.01 | 85.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PSPNetPre-training=COCO, FPS=11, #Param=67.6M2018.11 | 85.4 | — | — | 72.7 | 95 | 78.9 | 84.4 | 95.7 | 93.1 | 95.2 | 43.1 | 91.3 | 81.3 | 96.3 | 92.3 | 90.1 | 91.5 | 74.4 | 90.6 | 66.9 | 88.8 | 82 | — | 95.8 | — | — | — | — | — | — | — | — | — | — | — | |
| PSPNetPre-trained=MS-COCO2020.11 | 85.4 | — | — | 72.7 | 95 | 78.9 | 84.4 | 94.7 | 92 | 95.7 | 43.1 | 91 | 80.3 | 91.3 | 96.3 | 92.3 | 90.1 | 71.5 | 94.4 | 66.9 | 88.8 | 82 | — | 95.8 | — | — | — | — | — | — | — | — | — | — | — | |
| PSPNetpre-training=MS-COCO2020.11 | 85.4 | — | — | 72.7 | 95 | 78.9 | 84.4 | 94.7 | 92 | 95.7 | 43.1 | 91 | 80.3 | 91.3 | 96.3 | 92.3 | 90.1 | 71.5 | 94.4 | 66.9 | 88.8 | 82 | — | 95.8 | — | — | — | — | — | — | — | — | — | — | — | |
| PSPNetBackbone=ResNet-101, COCO pre-training=true2026.06 | 85.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SegNeXt-SBackbone=MSCAN-S2022.09 | 85.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SegNeXt-SBackbone=MSCAN-S, L. Params=13.9M2024.06 | 85.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ResNet38Training Data=Using VOC+COCO data2017.08 | 84.9 | — | — | 75.2 | 95.4 | 74.4 | 81.7 | 93.7 | 89.9 | 92.5 | 48.2 | 92 | 79.9 | 90.1 | 95.5 | 91.8 | 91.2 | 73 | 90.5 | 65.4 | 88.7 | 80.6 | — | 96.2 | — | — | — | — | — | — | — | — | — | — | — | |
| ResNet-38_MS_COCOPre-training=MS-COCO2017.06 | 84.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ResNet-38Pre-trained on MS-COCO=true2018.04 | 84.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ResNet-38multi-scale=true, training data=COCO2018.04 | 84.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ResNet-38ImageNet Pretrained=true, COCO Pretrained=true2019.01 | 84.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ResNet38Pre-training=COCO, Backbone=ResNet-38, FPS=13, #Param=55.9M2018.11 | 84.9 | — | — | 75.2 | 95.4 | 74.4 | 81.7 | 95 | 92.6 | 92.6 | 48.2 | 92 | 91 | 95.5 | 91.8 | 91.2 | 91 | 73 | 90.5 | 65.4 | 88.7 | 80.6 | — | 96.2 | — | — | — | — | — | — | — | — | — | — | — | |
| ResNet38Pre-trained=MS-COCO2020.11 | 84.9 | — | — | 75.2 | 95.4 | 74.4 | 81.7 | 95.6 | 89.8 | 94.1 | 43.7 | 92 | 77.2 | 90.8 | 93.4 | 91.8 | 91.2 | 65.4 | 88.7 | 73 | 90.5 | 80.6 | — | 96.2 | — | — | — | — | — | — | — | — | — | — | — | |
| ResNet38pre-training=MS-COCO2020.11 | 84.9 | — | — | 75.2 | 95.4 | 74.4 | 81.7 | 95.6 | 89.8 | 94.1 | 43.7 | 92 | 77.2 | 90.8 | 93.4 | 88.6 | 91.8 | 73 | 90.5 | 65.4 | 88.7 | 80.6 | — | 96.2 | — | — | — | — | — | — | — | — | — | — | — | |
| OCRNetBackbone=HRNetV2-W482022.09 | 84.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RefineNetTraining Data=Using VOC+COCO data2017.08 | 84.2 | — | — | 73.2 | 93.5 | 78.1 | 84.8 | 95.6 | 89.8 | 94.1 | 43.7 | 92 | 77.2 | 90.8 | 93.4 | 88.6 | 88.1 | 70.1 | 92.9 | 64.3 | 87.7 | 78.8 | — | 95 | — | — | — | — | — | — | — | — | — | — | — | |
| Multipath-RefineNet2017.06 | 84.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RefineNetPre-trained on MS-COCO=true2018.04 | 84.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Multipath RefineNet2018.04 | 84.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RefineNetImageNet Pretrained=true, COCO Pretrained=true2019.01 | 84.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FDNetTraining Data=VOC Only, denseCRF=Excluded2019.05 | 84.2 | — | — | 79.9 | 88.6 | 76.1 | 79.5 | 96.7 | 91.4 | 95.6 | 40.1 | 93 | 71.5 | 93.4 | 95.7 | 91.1 | 89.2 | 69.4 | 93.3 | 68 | 88.3 | 76.8 | — | 95.5 | — | — | — | — | — | — | — | — | — | — | — | |
| ShelfNet101Pre-training=COCO, Backbone=ResNet-101, FPS=42, #Param=57.7M2018.11 | 84.2 | — | — | 91.2 | 93.9 | — | — | — | — | — | 35.3 | 90.1 | 73.2 | 92.3 | — | 64.5 | 88 | 77.5 | — | — | — | — | — | 96.3 | — | — | — | — | — | — | — | — | — | — | — | |
| RefineNetPre-trained=MS-COCO2020.11 | 84.2 | — | — | 73.2 | 93.5 | 78.1 | 84.8 | 93.7 | 89.9 | 92.5 | 48.2 | 90.5 | 73 | 90.5 | 93.5 | 88.6 | 88.1 | 70.1 | 92.9 | 64.3 | 87.7 | 78.8 | — | 95 | — | — | — | — | — | — | — | — | — | — | — | |
| RefineNetpre-training=MS-COCO2020.11 | 84.2 | — | — | 73.2 | 93.5 | 78.1 | 84.8 | 93.7 | 89.9 | 92.5 | 48.2 | 91.8 | 73 | 90.5 | 93.7 | 88.1 | 91.2 | 70.1 | 92.9 | 64.3 | 87.7 | 78.8 | — | 95 | — | — | — | — | — | — | — | — | — | — | — | |
| Auto-DeepLab-MImageNet Pretrained=false, COCO Pretrained=true2019.01 | 84.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Large_Kernel_Matters2017.06 | 83.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GCNPre-trained on MS-COCO=true2018.04 | 83.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Large Kernel Matters2018.04 | 83.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SDNTraining Data=Only using VOC data2017.08 | 83.5 | — | — | 73.9 | 94 | 74.1 | 76.1 | 96.7 | 89.9 | 96.2 | 44.1 | 92.6 | 72.3 | 91.2 | 94.1 | 89.2 | 89.7 | 71.2 | 93 | 59 | 88.4 | 76.5 | — | 96.2 | — | — | — | — | — | — | — | — | — | — | — | |
| SDNTraining Data=VOC Only, denseCRF=Excluded2019.05 | 83.5 | — | — | 73.9 | 94 | 74.1 | 76.1 | 96.7 | 89.9 | 96.2 | 44.1 | 92.6 | 72.3 | 91.2 | 94.1 | 89.2 | 89.7 | 71.2 | 93 | 59 | 88.4 | 76.5 | — | 96.2 | — | — | — | — | — | — | — | — | — | — | — | |
| RefineNetPre-training=COCO, Backbone=ResNet-152, FPS=14, #Param=118M2018.11 | 83.4 | — | — | 73.2 | 93.5 | 78.1 | 84.8 | 95.6 | 89.8 | 94.1 | 43.7 | 92 | 88.5 | 92 | 92 | 91 | 88.1 | 70.1 | 92.9 | 64.3 | 87.7 | 78.8 | — | 95 | — | — | — | — | — | — | — | — | — | — | — | |
| SANetCOCO Pre-training=false2019.09 | 83.2 | — | — | 65.9 | 95.4 | 72 | 80.5 | 93.5 | 86.8 | 94.5 | 40.5 | 93.3 | 74.6 | 94.1 | — | — | — | — | — | — | — | — | — | 95.1 | — | — | — | — | — | — | — | — | — | — | — | |
| ResNet-DUClayers=152, model ensemble=false, multiscale testing=false2017.02 | 83.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TuSimple2017.06 | 83.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DUCPre-trained on MS-COCO=true2018.04 | 83.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Tusimple2018.04 | 83.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SegmenterBackbone=ViT-S, Supervision=Fully-supervised, Pre-trained Model=Improved ViT2023.04 | 83.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |