Semantic Segmentation on PST900 (test)
89.4mIoUStitchFusion
Evaluation results
| Method | Links | ||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| StitchFusionBackbone=Swin-Large-22k, Additional Strategies=None2024.08 | 89.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| HAPNetBackbone=ConvNext-L, Additional Strategies=Used2024.08 | 89 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CRMBackbone=Swin-B2023.03 | 88 | 99.6 | 79.5 | 89.6 | 89 | 82.2 | — | — | — | — | — | — | — | — | — | — | — | — | |
| CRMBackbone=Swin-S2023.03 | 86.9 | 99.6 | 76.2 | 91 | 87.2 | 80.8 | — | — | — | — | — | — | — | — | — | — | — | — | |
| DPLNetBackbone=MiT-B5, Params (M)=7.152023.12 | 86.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CACFNetBackbone=ConvNeXt-B, Params (M)=198.62023.12 | 86.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CRMBackbone=Swin-T2023.03 | 85.9 | 99.5 | 79.1 | 86 | 86.2 | 78.7 | — | — | — | — | — | — | — | — | — | — | — | — | |
| EAEF2023.03 | 85.56 | — | — | — | — | — | 91.42 | — | — | — | — | — | — | — | — | — | — | — | |
| FEANetBackbone=not specified, Additional Strategies=None2024.08 | 85.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| EAEFNetBackbone=not specified, Additional Strategies=Used2024.08 | 85.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| EGFNet-ConvNextBackbone=ConvNeXt-B2023.12 | 85.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CAINetInput Modality=RGB-T2024.01 | 84.74 | 99.5 | 77.21 | 88.02 | 80.3 | 78.69 | 94.27 | 99.66 | 95.87 | 96.09 | 88.38 | 91.35 | — | — | — | — | — | — | |
| StitchFusionBackbone=MiT-B4 (MiT-B3), Additional Strategies=None2024.08 | 84.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| LASNetInput Modality=RGB-T2024.01 | 84.4 | 99.46 | 77.75 | 86.48 | 82.8 | 75.49 | 91.63 | 99.77 | 91.81 | 90.8 | 92.36 | 83.43 | — | — | — | — | — | — | |
| ConvNeXt-TPretrain=CSP-42026.04 | 84.22 | 99.48 | 80.84 | 86.04 | 80.24 | 74.52 | — | — | — | — | — | — | — | — | — | — | — | — | |
| GMNet2023.03 | 84.12 | — | — | — | — | — | 89.61 | — | — | — | — | — | — | — | — | — | — | — | |
| GMNet21Input Modality=RGB-T2024.01 | 84.12 | 99.44 | 73.79 | 83.82 | 85.17 | 78.36 | 89.61 | 99.81 | 90.29 | 89.01 | 88.28 | 80.86 | — | — | — | — | — | — | |
| GMNetBackbone=ResNet-50, Params (M)=149.82023.12 | 84.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GMNet2023.03 | 84.1 | 99.4 | 73.8 | 83.8 | 85.2 | 78.4 | — | — | — | — | — | — | — | — | — | — | — | — | |
| StitchFusionBackbone=Swin-Tiny-1k (-22k), Additional Strategies=None2024.08 | 83.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Swin-TPretrain=CSP-42026.04 | 83.11 | 99.4 | 79.14 | 84.21 | 84.36 | 68.45 | — | — | — | — | — | — | — | — | — | — | — | — | |
| DSGBINet22Input Modality=RGB-T2024.01 | 82.87 | 99.39 | 79.31 | 85.11 | 74.99 | 75.56 | 91.81 | 99.73 | 94.53 | 88.65 | 94.78 | 81.37 | — | — | — | — | — | — | |
| CCFFNet5022Input Modality=RGB-T2024.01 | 82.1 | 99.4 | 79.9 | 75.8 | 82.8 | 72.7 | 86.9 | 99.9 | 89.7 | 77.5 | 87.6 | 72.7 | — | — | — | — | — | — | |
| VPFNetType=RGBT2023.07 | 81.41 | 99.36 | 77.66 | 81.99 | 73.14 | 74.88 | 91.46 | 99.7 | 93.45 | 87.48 | 92.74 | 83.91 | — | — | — | — | — | — | |
| GEBNetBackbone=ConvNeXt-S2023.12 | 81.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SA-Gate2023.03 | 79.05 | — | — | — | — | — | 84.71 | — | — | — | — | — | — | — | — | — | — | — | |
| SA-GateType=RGBD2023.07 | 79.05 | 99.25 | 72.97 | 79.77 | 81.01 | 62.22 | 84.71 | 99.74 | 89.88 | 89.03 | 80.7 | 64.19 | — | — | — | — | — | — | |
| MFFENet21Input Modality=RGB-T2024.01 | 78.98 | 99.4 | 66.38 | 81.02 | 72.5 | 75.6 | — | — | — | — | — | — | — | — | — | — | — | — | |
| MTANetBackbone=ResNet-152, Params (M)=121.92023.12 | 78.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MTANet22Input Modality=RGB-T2024.01 | 78.6 | 99.33 | 64.95 | 87.5 | 62.05 | 79.14 | — | — | — | — | — | — | — | — | — | — | — | — | |
| EGFNetType=RGBT2023.07 | 78.51 | 99.26 | 71.29 | 83.05 | 64.67 | 74.3 | 94.02 | 99.48 | 97.99 | 94.17 | 95.17 | 83.3 | — | — | — | — | — | — | |
| EGFNet22Input Modality=RGB-T2024.01 | 78.51 | 99.26 | 71.29 | 83.05 | 64.67 | 74.3 | 94.02 | 99.48 | 97.99 | 94.17 | 95.17 | 83.3 | — | — | — | — | — | — | |
| EGFNetBackbone=not specified, Additional Strategies=None2024.08 | 78.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| EGFNetBackbone=ResNet-152, Params (M)=201.32023.12 | 78.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FDCNet22Input Modality=RGB-T2024.01 | 77.11 | 99.15 | 71.52 | 72.17 | 70.36 | 72.36 | 85.96 | 99.72 | 82.52 | 77.45 | 91.77 | 78.36 | — | — | — | — | — | — | |
| RTFDNet (MiT-B2)Backbone=MiT-B2 [33], Parameters (M)=120.02026.03 | 76.88 | — | — | — | — | — | 81.19 | — | — | — | — | — | 90.28 | 84.3 | 65.47 | 62.2 | 87.81 | 84.15 | |
| MDRNet+Type=RGBT2023.07 | 74.62 | 99.07 | 63.47 | 76.27 | 63.04 | 71.26 | 91.04 | 99.3 | 90.17 | 93.53 | 86.63 | 85.56 | — | — | — | — | — | — | |
| ConvNeXt-TPretrain=IM-RGB2026.04 | 73.2 | 99.22 | 66.68 | 75.16 | 60.09 | 64.85 | — | — | — | — | — | — | — | — | — | — | — | — | |
| ACNet2023.03 | 71.81 | — | — | — | — | — | 78.67 | — | — | — | — | — | — | — | — | — | — | — | |
| ACNetType=RGBD2023.07 | 71.81 | 99.25 | 59.95 | 83.19 | 51.46 | 65.19 | 78.67 | 99.83 | 53.59 | 85.56 | 84.88 | 69.1 | — | — | — | — | — | — | |
| ABMDRNetType=RGBT2023.07 | 71.33 | 99 | 66.22 | 67.91 | 61.52 | 62.02 | 79.06 | 99.78 | 81.24 | 71.41 | 76.49 | 66.4 | — | — | — | — | — | — | |
| ABMDRNet2023.03 | 71.3 | 99 | 66.2 | 67.9 | 61.5 | 62 | — | — | — | — | — | — | — | — | — | — | — | — | |
| DANet-4cType=RGB2023.07 | 68.86 | 99.02 | 48.05 | 78.49 | 68.28 | 50.46 | 75.74 | 99.74 | 82.8 | 83.28 | 58.16 | 54.73 | — | — | — | — | — | — | |
| CRMBackbone=Swin-T [42], Parameters (M)=117.682026.03 | 68.74 | — | — | — | — | — | 71.24 | — | — | — | — | — | 88.91 | 85.22 | 35.87 | 35.1 | 88.93 | 85.89 | |
| PSTNetBackbone=ResNet-18, Params (M)=105.82023.12 | 68.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PSTNet2023.03 | 68.4 | 98.9 | 70.1 | 69.2 | 53.6 | 50 | — | — | — | — | — | — | — | — | — | — | — | — | |
| OursVariant=Full, Mode=RGB-T, Inference Latency (ms)=422019.09 | 68.36 | 98.85 | 70.12 | 69.2 | 53.6 | 50.03 | — | — | — | — | — | — | — | — | — | — | — | — | |
| PSTNet2023.03 | 68.36 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| PSTNetType=RGBT2023.07 | 68.36 | 98.85 | 70.12 | 69.2 | 53.6 | 50.03 | — | — | — | — | — | — | — | — | — | — | — | — | |
| PSTNet20Input Modality=RGB-T2024.01 | 68.36 | 98.85 | 70.12 | 69.2 | 53.6 | 50.03 | — | — | — | — | — | — | — | — | — | — | — | — | |
| OursVariant=RGB Stream, Mode=RGB, Inference Latency (ms)=182019.09 | 67.65 | 98.83 | 68.14 | 69.9 | 51.51 | 49.89 | — | — | — | — | — | — | — | — | — | — | — | — | |
| ABMDRNetBackbone=not specified, Additional Strategies=None2024.08 | 67.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| CCNet-4cType=RGB2023.07 | 67.14 | 98.95 | 49.24 | 78.82 | 55.93 | 52.77 | 76.6 | 99.62 | 74.62 | 86.09 | 65.15 | 57.51 | — | — | — | — | — | — | |
| Swin-TPretrain=IM-RGB2026.04 | 66.47 | 99.02 | 63.7 | 73.55 | 43.9 | 52.19 | — | — | — | — | — | — | — | — | — | — | — | — | |
| CCNet2023.03 | 66 | — | — | — | — | — | 73.43 | — | — | — | — | — | — | — | — | — | — | — | |
| RTFNet-50Type=RGBT2023.07 | 65.52 | 98.89 | 54.46 | 67.91 | 52.24 | 54.11 | 72.97 | 99.74 | 64.07 | 73.08 | 69.78 | 58.19 | — | — | — | — | — | — | |
| EAEFNetBackbone=ResNet-50 [40], Parameters (M)=200.42026.03 | 63.21 | — | — | — | — | — | 67.05 | — | — | — | — | — | 88.52 | 83.45 | 21.22 | 20.63 | 91.42 | 85.56 | |
| ERFNetMode=RGB-T, Inference Latency (ms)=312019.09 | 62.55 | 98.73 | 58.79 | 68.08 | 52.76 | 34.38 | — | — | — | — | — | — | — | — | — | — | — | — | |
| OursVariant=Naïve, Mode=RGB-T, Inference Latency (ms)=202019.09 | 62.23 | 98.52 | 62.53 | 59.61 | 47.08 | 43.39 | — | — | — | — | — | — | — | — | — | — | — | — | |
| StitchFusionBackbone=Swin-T [42]2026.03 | 62.06 | — | — | — | — | — | 65.87 | — | — | — | — | — | 87.17 | 82.11 | 20 | 19.44 | 90.43 | 84.64 | |
| ERFNetMode=RGB, Inference Latency (ms)=302019.09 | 61.85 | 98.69 | 61.18 | 65.28 | 42.4 | 41.69 | — | — | — | — | — | — | — | — | — | — | — | — | |
| RTFNetBackbone=ResNet-152, Params (M)=337.12023.12 | 60.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RTFNet2023.03 | 60.46 | — | — | — | — | — | 65.69 | — | — | — | — | — | — | — | — | — | — | — | |
| RTFNetBackbone=RTFNet-152, Mode=RGB-T, Inference Latency (ms)=1272019.09 | 57.61 | 98.92 | 52.03 | 75.3 | 25.37 | 36.43 | — | — | — | — | — | — | — | — | — | — | — | — | |
| RTFNetBackbone=not specified, Additional Strategies=None2024.08 | 57.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| RTFNet-152Backbone=ResNet-1522023.03 | 57.6 | 98.9 | 52 | 75.3 | 25.4 | 36.4 | — | — | — | — | — | — | — | — | — | — | — | — | |
| Efficient FCN2023.03 | 57.27 | — | — | — | — | — | 66.75 | — | — | — | — | — | — | — | — | — | — | — | |
| MFNetMode=RGB-T, Inference Latency (ms)=232019.09 | 57.02 | 98.63 | 60.35 | 64.27 | 41.13 | 20.7 | — | — | — | — | — | — | — | — | — | — | — | — | |
| MFNet17Input Modality=RGB-T2024.01 | 57.02 | 98.63 | 60.35 | 64.27 | 41.13 | 20.7 | — | — | — | — | — | — | — | — | — | — | — | — | |
| MFNetBackbone=not specified, Additional Strategies=None2024.08 | 57 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MFNet2023.03 | 57 | 98.6 | 60.4 | 64.3 | 41.1 | 20.7 | — | — | — | — | — | — | — | — | — | — | — | — | |
| UNetMode=RGB, Inference Latency (ms)=122019.09 | 54.99 | 98.43 | 49.28 | 63.64 | 40.26 | 23.37 | — | — | — | — | — | — | — | — | — | — | — | — | |
| CAINetBackbone=mobilenet-v2, Additional Strategies=None2024.08 | 54.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| UNetMode=RGB-T, Inference Latency (ms)=122019.09 | 52.74 | 97.95 | 42.96 | 52.89 | 38.27 | 31.64 | — | — | — | — | — | — | — | — | — | — | — | — | |
| MFNetType=RGBT2023.07 | 50.34 | 97.85 | 37.31 | 63.02 | 4.87 | 8.63 | 63.5 | 98.89 | 68.59 | 77.49 | 57.15 | 15.4 | — | — | — | — | — | — | |
| RTFNetBackbone=RTFNet-50, Mode=RGB-T, Inference Latency (ms)=422019.09 | 48.4 | 98.84 | 43.49 | 70.58 | 1 | 28 | — | — | — | — | — | — | — | — | — | — | — | — | |
| Fast-SCNNMode=RGB, Inference Latency (ms)=182019.09 | 48.22 | 98.57 | 34.54 | 66.79 | 20.63 | 20.53 | — | — | — | — | — | — | — | — | — | — | — | — | |
| MAVNetMode=RGB-T, Inference Latency (ms)=172019.09 | 47.74 | 97.89 | 22.58 | 51.52 | 31.94 | 34.73 | — | — | — | — | — | — | — | — | — | — | — | — | |
| Fast-SCNNMode=RGB-T, Inference Latency (ms)=182019.09 | 47.15 | 98.51 | 35.48 | 64.6 | 15.5 | 21.68 | — | — | — | — | — | — | — | — | — | — | — | — | |
| MAVNetMode=RGB, Inference Latency (ms)=162019.09 | 45.51 | 98.22 | 28.31 | 58.5 | 33.67 | 9.01 | — | — | — | — | — | — | — | — | — | — | — | — |