Semantic Segmentation on WeedMap (test)
76.9mIoUFCBNet-large
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| FCBNet-largeModality=RGB-NIR-RE, Inference time (s)=0.0257, Training time (h)=0.215, Total params. (M)=202.329, Trainable params. (M)=4.555, GFLOPS=13.7842026.02 | 76.9 | 98.7 | 55.1 | |
| U-NetModality=RGB-NIR-RE, Inference time (s)=0.0117, Training time (h)=0.131, Total params. (M)=32.528, Trainable params. (M)=32.528, GFLOPS=30.3622026.02 | 76.8 | 98.6 | 55 | |
| FCBNet-largeModality=RGB-NIR, Inference time (s)=0.0251, Training time (h)=0.211, Total params. (M)=202.325, Trainable params. (M)=4.555, GFLOPS=13.7492026.02 | 76.7 | 98.6 | 54.8 | |
| FCBNet-largeModality=RGB, Inference time (s)=0.0252, Training time (h)=0.201, Total params. (M)=202.322, Trainable params. (M)=4.555, GFLOPS=13.7142026.02 | 76.6 | 98.7 | 54.6 | |
| FCBNet-tinyModality=RGB-NIR-RE, Inference time (s)=0.0107, Training time (h)=0.098, Total params. (M)=30.607, Trainable params. (M)=2.015, GFLOPS=9.2822026.02 | 76.4 | 98.6 | 54.3 | |
| FCBNet-tinyModality=RGB-NIR, Inference time (s)=0.0103, Training time (h)=0.091, Total params. (M)=30.605, Trainable params. (M)=2.015, GFLOPS=9.2642026.02 | 76.3 | 98.6 | 54 | |
| SK-U-NetModality=RGB-NIR-RE, Inference time (s)=0.0140, Training time (h)=0.178, Total params. (M)=34.450, Trainable params. (M)=34.450, GFLOPS=31.7512026.02 | 76.2 | 98.4 | 53.9 | |
| FCBNet-tinyModality=RGB, Inference time (s)=0.0095, Training time (h)=0.083, Total params. (M)=30.604, Trainable params. (M)=2.015, GFLOPS=9.2472026.02 | 76.1 | 98.5 | 53.7 | |
| SK-U-NetModality=RGB-NIR, Inference time (s)=0.0164, Training time (h)=0.170, Total params. (M)=34.447, Trainable params. (M)=34.447, GFLOPS=31.6062026.02 | 76.1 | 98.2 | 53.9 | |
| DeepLabV3+Modality=RGB-NIR-RE, Inference time (s)=0.0200, Training time (h)=0.189, Total params. (M)=26.684, Trainable params. (M)=26.684, GFLOPS=26.1372026.02 | 76 | 98.6 | 53.4 | |
| SegFormerModality=RGB-NIR-RE, Inference time (s)=0.0189, Training time (h)=0.421, Total params. (M)=84.601, Trainable params. (M)=84.601, GFLOPS=34.1892026.02 | 76 | 98.5 | 53.5 | |
| SegFormerModality=RGB-NIR, Inference time (s)=0.0180, Training time (h)=0.421, Total params. (M)=84.598, Trainable params. (M)=84.598, GFLOPS=34.1532026.02 | 75.8 | 98.5 | 53.2 | |
| SK-U-NetModality=RGB, Inference time (s)=0.0140, Training time (h)=0.160, Total params. (M)=34.444, Trainable params. (M)=34.444, GFLOPS=31.4612026.02 | 75.8 | 98.3 | 53.2 | |
| WeedSenseModality=RGB-NIR-RE, Inference time (s)=0.0196, Training time (h)=0.166, Total params. (M)=12.724, Trainable params. (M)=12.724, GFLOPS=9.0182026.02 | 75.2 | 98.4 | 51.9 | |
| WeedSenseModality=RGB-NIR, Inference time (s)=0.0190, Training time (h)=0.151, Total params. (M)=12.723, Trainable params. (M)=12.723, GFLOPS=8.9852026.02 | 74.9 | 98.4 | 51.5 | |
| DeepLabV3+Modality=RGB-NIR, Inference time (s)=0.0190, Training time (h)=0.170, Total params. (M)=26.681, Trainable params. (M)=26.681, GFLOPS=25.9932026.02 | 74.9 | 98.4 | 51.5 | |
| DeepLabV3+Modality=RGB, Inference time (s)=0.0190, Training time (h)=0.161, Total params. (M)=26.678, Trainable params. (M)=26.678, GFLOPS=25.8492026.02 | 74.6 | 98.3 | 50.9 | |
| SegFormerModality=RGB, Inference time (s)=0.0177, Training time (h)=0.420, Total params. (M)=84.595, Trainable params. (M)=84.595, GFLOPS=34.1162026.02 | 74.6 | 98.6 | 50.6 | |
| U-NetModality=RGB-NIR, Inference time (s)=0.0112, Training time (h)=0.126, Total params. (M)=32.524, Trainable params. (M)=32.524, GFLOPS=30.2182026.02 | 74.5 | 98.1 | 50.8 | |
| WeedSenseModality=RGB, Inference time (s)=0.0115, Training time (h)=0.138, Total params. (M)=12.721, Trainable params. (M)=12.721, GFLOPS=8.9522026.02 | 72.5 | 97.9 | 47.1 | |
| U-NetModality=RGB, Inference time (s)=0.0109, Training time (h)=0.119, Total params. (M)=32.521, Trainable params. (M)=32.521, GFLOPS=30.0732026.02 | 70.1 | 97.5 | 42.7 | |
| TransferFSSBackbone=Swin-B, Number of support images (K)=502025.12 | 64.18 | — | — | |
| DistillFSSBackbone=Swin-B, Number of support images (K)=502025.12 | 61.96 | — | — | |
| DistillFSSBackbone=Swin-B, Number of support images (K)=252025.12 | 59.43 | — | — | |
| TransferFSSBackbone=Swin-B, Number of support images (K)=102025.12 | 58.65 | — | — | |
| TransferFSSBackbone=Swin-B, Number of support images (K)=252025.12 | 57.55 | — | — | |
| DistillFSSBackbone=Swin-B, Number of support images (K)=102025.12 | 55.38 | — | — | |
| DistillFSSBackbone=ResNet-50, Number of support images (K)=502025.12 | 54.48 | — | — | |
| TransferFSSBackbone=ResNet-50, Number of support images (K)=502025.12 | 54.03 | — | — | |
| TransferFSSBackbone=ResNet-50, Number of support images (K)=252025.12 | 52.09 | — | — | |
| TransferFSSBackbone=Swin-B, Number of support images (K)=52025.12 | 51.01 | — | — | |
| DistillFSSBackbone=ResNet-50, Number of support images (K)=252025.12 | 48.38 | — | — | |
| DistillFSSBackbone=ResNet-50, Number of support images (K)=102025.12 | 47.99 | — | — | |
| TransferFSSBackbone=ResNet-50, Number of support images (K)=102025.12 | 46.18 | — | — | |
| DistillFSSBackbone=Swin-B, Number of support images (K)=52025.12 | 44.43 | — | — | |
| DistillFSSBackbone=ResNet-50, Number of support images (K)=52025.12 | 32.4 | — | — | |
| TransferFSSBackbone=ResNet-50, Number of support images (K)=52025.12 | 31.73 | — | — | |
| PATNetBackbone=ResNet-50, Number of support images (K)=502025.12 | 6.96 | — | — | |
| PATNetBackbone=ResNet-50, Number of support images (K)=252025.12 | 6.95 | — | — | |
| BAMBackbone=ResNet-50, Number of support images (K)=52025.12 | 6.63 | — | — | |
| PATNetBackbone=ResNet-50, Number of support images (K)=52025.12 | 6.55 | — | — | |
| BAMBackbone=ResNet-50, Number of support images (K)=252025.12 | 6.36 | — | — | |
| PATNetBackbone=ResNet-50, Number of support images (K)=102025.12 | 6.3 | — | — | |
| BAMBackbone=ResNet-50, Number of support images (K)=502025.12 | 6.16 | — | — | |
| DCAMABackbone=Swin-B, Number of support images (K)=102025.12 | 6.13 | — | — | |
| BAMBackbone=ResNet-50, Number of support images (K)=102025.12 | 5.53 | — | — | |
| DCAMABackbone=ResNet-50, Number of support images (K)=102025.12 | 5.11 | — | — | |
| DCAMABackbone=ResNet-50, Number of support images (K)=52025.12 | 5.1 | — | — | |
| DCAMABackbone=ResNet-50, Number of support images (K)=502025.12 | 5.01 | — | — | |
| DCAMABackbone=ResNet-50, Number of support images (K)=252025.12 | 4.9 | — | — | |
| DCAMABackbone=Swin-B, Number of support images (K)=52025.12 | 4.7 | — | — | |
| DCAMABackbone=Swin-B, Number of support images (K)=252025.12 | 4.38 | — | — | |
| HDMNetBackbone=ResNet-50, Number of support images (K)=502025.12 | 4.04 | — | — | |
| HDMNetBackbone=ResNet-50, Number of support images (K)=252025.12 | 3.81 | — | — | |
| DCAMABackbone=Swin-B, Number of support images (K)=502025.12 | 3.78 | — | — | |
| LabelAnythingBackbone=ViT-B, Number of support images (K)=502025.12 | 3.74 | — | — | |
| HDMNetBackbone=ResNet-50, Number of support images (K)=102025.12 | 3.71 | — | — | |
| LabelAnythingBackbone=ViT-B, Number of support images (K)=252025.12 | 3.69 | — | — | |
| DMTNetBackbone=ResNet-50, Number of support images (K)=52025.12 | 2.61 | — | — | |
| DMTNetBackbone=ResNet-50, Number of support images (K)=252025.12 | 2.4 | — | — | |
| DMTNetBackbone=ResNet-50, Number of support images (K)=502025.12 | 2.3 | — | — | |
| LabelAnythingBackbone=ViT-B, Number of support images (K)=52025.12 | 2.28 | — | — | |
| LabelAnythingBackbone=ViT-B, Number of support images (K)=102025.12 | 2.21 | — | — | |
| DMTNetBackbone=ResNet-50, Number of support images (K)=102025.12 | 2.17 | — | — | |
| HDMNetBackbone=ResNet-50, Number of support images (K)=52025.12 | 1.24 | — | — |