Semantic Segmentation on NYUDv2 40-class (test)
53.5mIoUPDCNet
Evaluation results
| Method | Links | ||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| PDCNetBackbone=ResNet101, Information Source=Depth, Multi-scale=true2023.02 | 53.5 | 78.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| InverseFormBackbone=Transformer, Information Source=Depth, Multi-scale=true2023.02 | 53.1 | 78.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PDCNetBackbone=ResNet101, Information Source=Depth, Multi-scale=false2023.02 | 52.7 | 77.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ADSDCategory=encoder-decoder, Data=RGBD, Backbone=ResNet-50, Scale=single-scale inference2022.01 | 52.5 | 77.5 | 65.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ADSDBackbone=ResNet50, Information Source=Depth, Multi-scale=false2023.02 | 52.5 | 77.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SA-GateCategory=encoder-decoder, Data=RGBD, Backbone=ResNet-101, Scale=multi-scale inference2022.01 | 52.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SA-GateBackbone=ResNet101, Information Source=Depth, Multi-scale=true2023.02 | 52.4 | 77.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ShapeConvBackbone=ResNet101, Information Source=HHA, Multi-scale=true2023.02 | 51.3 | 76.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ShapeConvMulti-scale testing=true2021.08 | 51.3 | 76.4 | 63.5 | 63 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CANetCategory=encoder-decoder, Data=RGBD, Backbone=ResNet-101, Scale=single-scale inference2022.01 | 51.2 | 76.6 | 63.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SGNetBackbone=ResNet50, Information Source=Depth, Multi-scale=true2023.02 | 51.1 | 76.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SGNetMulti-scale testing=true2021.08 | 51.1 | 76.8 | 63.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PSDCategory=multi-task encoder-decoder, Data=RGB, Backbone=ResNet-50, Scale=single-scale inference2022.01 | 51 | 77 | 58.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Malleable 2.5DCategory=encoder-decoder, Data=RGBD, Backbone=ResNet-101, Scale=multi-scale inference2022.01 | 50.9 | 76.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| M2.5DBackbone=ResNet101, Information Source=Depth, Multi-scale=true2023.02 | 50.9 | 76.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| M2.5DMulti-scale testing=true2021.08 | 50.9 | 76.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ESANetCategory=encoder-decoder, Data=RGBD, Backbone=ResNet-50, Scale=single-scale inference2022.01 | 50.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PAPCategory=multi-task encoder-decoder, Data=RGB, Backbone=ResNet-50, Scale=single-scale inference2022.01 | 50.4 | 76.2 | 62.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SGNetCategory=encoder-decoder, Data=RGBD, Backbone=ResNet-101, Scale=single-scale inference2022.01 | 50.3 | 76.4 | 62.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PAD-NetCategory=multi-task encoder-decoder, Data=RGB, Backbone=ResNet-50, Scale=single-scale inference2022.01 | 50.2 | 75.2 | 62.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ShapeConvBackbone=ResNet101, Information Source=HHA, Multi-scale=false2023.02 | 50.2 | 75.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ShapeConvMulti-scale testing=false2021.08 | 50.2 | 75.8 | 62.8 | 62.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RDFNet-152Category=encoder-decoder, Data=RGB, Backbone=ResNet-152, Scale=multi-scale inference2022.01 | 50.1 | 76 | 62.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RDFBackbone=ResNet152, Information Source=HHA, Multi-scale=true2023.02 | 50.1 | 76 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RDFMulti-scale testing=true2021.08 | 50.1 | 76 | 62.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MTI-NetCategory=multi-task encoder-decoder, Data=RGB, Backbone=HRNet48-V2, Scale=single-scale inference2022.01 | 49 | 75.3 | 62.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| D-CNNBackbone=VGG16, Information Source=Depth, Multi-scale=false2023.02 | 48.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| D-CNNMulti-scale testing=false2021.08 | 48.4 | — | 61.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ACNetCategory=encoder-decoder, Data=RGBD, Backbone=ResNet-50, Scale=single-scale inference2022.01 | 48.3 | — | 63.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ACNetBackbone=ResNet50, Information Source=Depth, Multi-scale=false2023.02 | 48.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| ACNetMulti-scale testing=false2021.08 | 48.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CFNMulti-scale testing=true2021.08 | 47.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RefineNetCategory=atrous/dilated convolution, Data=RGB, Backbone=ResNet-152, Scale=multi-scale inference2022.01 | 47.6 | 74.4 | 59.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FDNet-16sInput modality=RGB only, Multi-scale inference=true2019.05 | 47.4 | 73.9 | 60.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RedNetCategory=encoder-decoder, Data=RGBD, Backbone=ResNet-50, Scale=single-scale inference2022.01 | 47.2 | — | 62.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RefineNet-Res152training data=RGB, backbone=ResNet-1522016.11 | 46.5 | 73.6 | 58.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RefineNetInput modality=RGB only, Multi-scale inference=true2019.05 | 46.5 | 73.6 | 58.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| TRLCategory=multi-task encoder-decoder, Data=RGB, Backbone=ResNet-50, Scale=single-scale inference2022.01 | 46.4 | 76.2 | 56.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| LSD-GFCategory=encoder-decoder, Data=RGBD, Backbone=VGG-16, Scale=single-scale inference2022.01 | 45.9 | 71.9 | 60.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| LSD-GFBackbone=VGG16, Information Source=HHA, Multi-scale=false2023.02 | 45.9 | 71.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| LSD-GFMulti-scale testing=false2021.08 | 45.9 | 71.9 | 60.7 | 59.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MMAF-NetBackbone=ResNet101, Information Source=Depth, Multi-scale=false2023.02 | 44.8 | 72.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MMAF-NetMulti-scale testing=false2021.08 | 44.8 | 72.2 | 59.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PU-LoopCategory=multi-task encoder-decoder, Data=RGB, Backbone=ResNet-50, Scale=single-scale inference2022.01 | 44.5 | 72.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| D-CNNCategory=atrous/dilated convolution, Data=RGBD, Backbone=VGG-16, Scale=single-scale inference2022.01 | 43.9 | — | 56.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| 3DGNNCategory=atrous/dilated convolution, Data=RGBD, Backbone=VGG-16, Scale=multi-scale inference2022.01 | 43.1 | — | 55.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| 3DGNNMulti-scale testing=true2021.08 | 43.1 | — | 55.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RTJ-AACategory=multi-task encoder-decoder, Data=RGB, Backbone=MobileNetV2, Scale=single-scale inference2022.01 | 42 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CFNCategory=atrous/dilated convolution, Data=RGBD, Backbone=VGG-16, Scale=multi-scale inference2022.01 | 41.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FeatMap-Net (Contextual Deep CRFs)training data=RGB2015.04 | 40.6 | 70 | 53.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Contexttraining data=RGB2016.11 | 40.6 | 70 | 53.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PiecewiseInput modality=RGB only, Multi-scale inference=true2019.05 | 40.6 | 70 | 53.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| STD2Pvariant=full model2016.04 | 40.1 | 70.1 | 53.8 | 55.7 | 72.7 | 85.7 | 55.4 | 73.6 | 58.5 | 60.1 | 42.7 | 30.2 | 42.1 | 41.9 | 52.9 | 59.7 | 46.7 | 13.5 | 9.4 | 40.7 | 44.1 | 42 | 34.5 | 35.6 | 22.2 | 55.9 | 29.8 | 41.7 | 52.5 | 21.1 | 34.4 | 15.5 | 7.8 | 29.2 | 60.7 | 42.2 | 62.7 | 47.4 | 38.6 | 28.5 | 23.5 | 7.3 | 18.8 | 15.1 | |
| DeepLab-LFOV2016.04 | 39.4 | 70.3 | 49.6 | 54.7 | 70.2 | 85.2 | 55.3 | 68.9 | 60.5 | 59.8 | 44.5 | 25.4 | 47.8 | 42.6 | 47.9 | 57.7 | 52.4 | 20.7 | 9.1 | 36 | 36.9 | 41.4 | 32.5 | 16 | 17.8 | 58.4 | 20.5 | 45.1 | 48 | 21 | 41.5 | 9.4 | 8 | 14.3 | 67 | 41.8 | 69.7 | 46.8 | 40.1 | 45.1 | 2.1 | 20.7 | 12.4 | 33.5 | |
| DeepLab-LFOVCategory=atrous/dilated convolution, Data=RGBD, Backbone=VGG-16, Scale=multi-scale inference2022.01 | 39.4 | 70.3 | 49.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLab-LFOVBackbone=VGG16, Information Source=Depth, Multi-scale=true2023.02 | 39.4 | 70.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FeatMap-Net (Baseline + sliding pyramid pooling + multi-scales + boundary refinement)training data=RGB2015.04 | 38.3 | 68.5 | 50.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| STD2Pvariant=superpixel+2016.04 | 38.1 | 68.4 | 52.1 | 54 | 72.4 | 84.3 | 52 | 71.5 | 54.3 | 58.8 | 37.9 | 28.2 | 41.9 | 38.5 | 52.3 | 58.2 | 49.7 | 14.3 | 8.1 | 42.9 | 35.9 | 40.8 | 27.7 | 31.9 | 19.3 | 55.6 | 28.2 | 38.3 | 46.9 | 17.6 | 31.2 | 11 | 6.5 | 28.2 | 66.7 | 34.1 | 62.8 | 47.8 | 35.1 | 26.4 | 8.8 | 19.3 | 10.9 | 29.2 | |
| FeatMap-Net (Baseline + sliding pyramid pooling + multi-scales)training data=RGB2015.04 | 37 | 67 | 50.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| DeepLab2016.04 | 36.8 | 68.7 | 46.9 | 52.5 | 67.9 | 83 | 53.1 | 66.8 | 57.8 | 57.8 | 43.4 | 19.4 | 45.5 | 41.5 | 49.3 | 58.3 | 47.8 | 15.5 | 7.3 | 32.9 | 34.3 | 40.2 | 23.7 | 18.7 | 20.2 | 55.1 | 22.1 | 30.6 | 49.4 | 21.8 | 32.1 | 6.4 | 5.8 | 14.8 | 55.3 | 37.7 | 57.9 | 47.7 | 40 | 44.7 | 6.6 | 18 | 12.9 | 36.8 | |
| DeepLabCategory=atrous/dilated convolution, Data=RGBD, Backbone=VGG-16, Scale=multi-scale inference2022.01 | 36.8 | 68.7 | 46.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| STD2Pvariant=superpixel2016.04 | 36 | 68.5 | 48.7 | 52.9 | 70.9 | 83.4 | 52.6 | 68.5 | 54.1 | 56 | 40.4 | 25.5 | 38.4 | 40.9 | 51.5 | 54.8 | 47.3 | 11.3 | 7.5 | 34.5 | 41.6 | 37.7 | 20.1 | 15.9 | 20.1 | 56.8 | 28.8 | 23.8 | 51.8 | 19.1 | 26.6 | 29.3 | 6.8 | 4.7 | 66.1 | 37.4 | 56.1 | 46.3 | 34.5 | 26.7 | 5.8 | 12.7 | 12.1 | 30.6 | |
| CRF-RNN2016.04 | 35.4 | 66.3 | 48.9 | 51 | 70.3 | 81.5 | 49.6 | 64.6 | 51.4 | 50.6 | 35.9 | 24.6 | 38.1 | 36 | 48.8 | 52.6 | 47.6 | 13.2 | 7.6 | 34.8 | 33.2 | 34.7 | 20.8 | 24 | 15 | 60.9 | 29.5 | 31.2 | 41.1 | 18.2 | 25.6 | 23 | 7.4 | 13.9 | 57.9 | 31.4 | 57.2 | 45.4 | 36.9 | 39.1 | 4.9 | 14.6 | 9.5 | 33.8 | |
| Multi-Scale CNN2016.04 | 34.1 | — | 45.1 | 51.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MS CNNCategory=multi-task encoder-decoder, Data=RGB, Backbone=VGG-16, Scale=single-scale inference2022.01 | 34.1 | 65.6 | 45.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCN-16sInput Modality=RGB-HHA, Architecture Stride=16s2014.11 | 34 | 65.4 | 46.1 | 49.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCN-HHAtraining data=RGB-D2015.04 | 34 | 65.4 | 46.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCN-HHAtraining data=RGB-D2016.11 | 34 | 65.4 | 46.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCN2016.04 | 34 | 65.4 | 46.1 | 49.5 | 69.9 | 79.4 | 50.3 | 66 | 47.5 | 53.2 | 32.8 | 22.1 | 39 | 36.1 | 50.5 | 54.2 | 45.8 | 11.9 | 8.6 | 32.5 | 31 | 37.5 | 22.4 | 13.6 | 18.3 | 59.1 | 27.3 | 27 | 41.9 | 15.9 | 26.1 | 14.1 | 6.5 | 12.9 | 57.6 | 30.1 | 61.3 | 44.8 | 32.1 | 39.2 | 4.8 | 15.2 | — | 30 | |
| FCN-HHAInput modality=RGB only, Multi-scale inference=true2019.05 | 34 | 65.4 | 46.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCNCategory=encoder-decoder, Data=RGBD, Backbone=VGG-16, Scale=single-scale inference2022.01 | 34 | 65.4 | 46.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCNBackbone=VGG16, Information Source=HHA, Multi-scale=false2023.02 | 34 | 65.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCNMulti-scale testing=false2021.08 | 34 | 65.4 | 46.1 | 49.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCN-32sInput modality=RGB-HHA, Fusion strategy=late fusion2016.05 | 33.3 | 65.3 | 44 | 48.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCN-32sInput Modality=RGB-HHA, Architecture Stride=32s2014.11 | 32.8 | 64.3 | 44.9 | 48 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FeatMap-Net (Baseline + sliding pyramid pooling)training data=RGB2015.04 | 32.4 | 63.5 | 45.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Bayesian SegNet2016.04 | 32.4 | — | 45.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Bayesian SegNetInput modality=RGB only, Multi-scale inference=true2019.05 | 32.4 | 68 | 45.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| B-SegNetCategory=encoder-decoder, Data=RGB, Backbone=VGG-16, Scale=single-scale inference2022.01 | 32.4 | 68 | 45.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCN-32sInput modality=RGB-D, Fusion strategy=early fusion2016.05 | 31.7 | 62.1 | 44.8 | 46.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCN-32sInput modality=RGB2016.05 | 31.6 | 61.8 | 44.7 | 46 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Mutex Constraints2016.04 | 31.5 | 63.8 | — | 48.5 | 65.6 | 79.2 | 51.9 | 66.7 | 41 | 55.7 | 36.5 | 20.3 | 33.2 | 32.6 | 44.6 | 53.6 | 49.1 | 10.8 | 9.1 | 47.6 | 27.6 | 42.5 | 30.2 | 32.7 | 12.6 | 56.7 | 8.9 | 21.6 | 19.2 | 28 | 28.6 | 22.9 | 1.6 | 1 | 9.6 | 30.6 | 48.4 | 41.8 | 28.1 | 27.6 | 0 | 9.8 | 7.6 | 24.5 | |
| E2S22016.04 | 31 | 58.1 | 52.9 | 44.2 | 56.9 | 67.8 | 50 | 59.5 | 43.8 | 44.3 | 31.3 | 24.6 | 37.9 | 32.7 | 46.1 | 45 | 51.8 | 15.8 | 9.1 | 38 | 34.8 | 31.5 | 31.7 | 25.3 | 14.2 | 39.7 | 26.7 | 27.1 | 35.2 | 17.8 | 21 | 19.9 | 7.4 | 36.9 | 35 | 17.6 | 31.8 | 36.3 | 14.8 | 26 | 9.9 | 14.5 | 9.3 | 20.9 | |
| FCN-32sInput Modality=RGBD, Architecture Stride=32s2014.11 | 30.5 | 61.5 | 42.4 | 45.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FeatMap-Net (FullyConvNet Baseline)training data=RGB2015.04 | 30.5 | 61.5 | 43.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCN-32sInput Modality=RGB, Architecture Stride=32s2014.11 | 29.2 | 60 | 42.2 | 43.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCN-32straining data=RGB2015.04 | 29.2 | 60 | 42.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCN-32straining data=RGB2016.11 | 29.2 | 60 | 42.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Gupta et al.2014.11 | 28.6 | 60.3 | — | 47 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Gupta et al.2016.05 | 28.6 | 60.3 | — | 47 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Gupta et al.training data=RGB-D2015.04 | 28.6 | 60.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Gupta et al.training data=RGB-D2016.11 | 28.6 | 60.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RGBD R-CNN2016.04 | 28.6 | 60.3 | — | 47 | 68 | 81.3 | 44.9 | 65 | 47.9 | 47.9 | 29.9 | 20.3 | 32.6 | 18.1 | 40.3 | 51.3 | 42 | 11.3 | 3.5 | 29.1 | 34.8 | 34.4 | 16.4 | 28 | 4.7 | 60.5 | 6.4 | 14.5 | 31 | 14.3 | 16.3 | 4.2 | 2.1 | 14.2 | 27.2 | 55.1 | 37.5 | 34.8 | 38.2 | 0.2 | 0.2 | 7.1 | 6.1 | 23.1 | |
| BIiterations=30002016.04 | 27.7 | 58.9 | 39.3 | 43 | 61.7 | 68.1 | 45.2 | 50.6 | 38.9 | 40.3 | 26.2 | 20.9 | 36 | 34.4 | 40.8 | 31.6 | 48.3 | 9.3 | 7.9 | 30.8 | 22.9 | 19.5 | 13.9 | 16.1 | 13.7 | 42.5 | 21.3 | 16.6 | 30.9 | 14.9 | 23.3 | 17.8 | 3.3 | 9.9 | 44.7 | 15.8 | 53.8 | 32.1 | 22.8 | 19 | 0.1 | 12.3 | 5.3 | 23.2 | |
| BI (3000)Category=atrous/dilated convolution, Data=RGBD, Backbone=VGG-16, Scale=single-scale inference2022.01 | 27.7 | 58.9 | 39.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| BIiterations=10002016.04 | 27.1 | 57.7 | 37.8 | 41.9 | 62.8 | 66.8 | 44.2 | 47.7 | 35.8 | 35.9 | 10.9 | 18.3 | 21.5 | 35.9 | 41.5 | 30.9 | 47.4 | 12.8 | 8.5 | 29.3 | 20.3 | 21.7 | 13 | 18.2 | 14.1 | 44.7 | 10.9 | 21.5 | 30.4 | 18.8 | 22.3 | 17.7 | 5.5 | 12.4 | 45.9 | 15.8 | 56.5 | 32.2 | 24.7 | 17.1 | 0.1 | 12.2 | 6.7 | 21.9 | |
| FCN-32sInput modality=HHA2016.05 | 25.2 | 58.3 | 35.7 | 41.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCN-32sInput Modality=HHA, Architecture Stride=32s2014.11 | 24.2 | 57.1 | 35.2 | 40.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SegNetInput modality=RGB only, Multi-scale inference=true2019.05 | 23.6 | 66.1 | 36 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |