Hypercorrelation Squeeze for Few-Shot Segmentation

Benchmark Model Rank Results
few-shot-semantic-segmentation-on-fss-1000-5-shotHSNet (ResNet-101)#15Mean IoU: 88.5
few-shot-semantic-segmentation-on-fss-1000-5-shotHSNet (ResNet-50)#19Mean IoU: 87.8
few-shot-semantic-segmentation-on-fss-1000-5-shotHSNet (VGG-16)#21Mean IoU: 85.8
few-shot-semantic-segmentation-on-fss-1000-1-shotHSNet (ResNet-101)#18Mean IoU: 86.5
few-shot-semantic-segmentation-on-fss-1000-1-shotHSNet (ResNet-50)#22Mean IoU: 85.5
few-shot-semantic-segmentation-on-fss-1000-1-shotHSNet (VGG-16)#24Mean IoU: 82.3
few-shot-semantic-segmentation-on-coco-20i-5-shotHSNet (ResNet-101)#35Mean IoU: 49.5FB-IoU: 72.4learnable parameters (million): 2.5
few-shot-semantic-segmentation-on-coco-20i-5-shotHSNet (ResNet-50)#50Mean IoU: 46.9FB-IoU: 70.7learnable parameters (million): 2.5
few-shot-semantic-segmentation-on-pascal-5i-5-shotHSNet (ResNet-101)#40Mean IoU: 70.4FB-IoU: 80.6learnable parameters (million): 2.5
few-shot-semantic-segmentation-on-pascal-5i-5-shotHSNet (ResNet-50)#46Mean IoU: 69.5FB-IoU: 80.6learnable parameters (million): 2.5
few-shot-semantic-segmentation-on-pascal-5i-5-shotHSNet (VGG-16)#74Mean IoU: 64.1FB-IoU: 76.6
few-shot-semantic-segmentation-on-pascal-5i-1-shotHSNet (ResNet-101)#42Mean IoU: 66.2FB-IoU: 77.6learnable parameters (million): 2.5
few-shot-semantic-segmentation-on-pascal-5i-1-shotHSNet (ResNet-50)#62Mean IoU: 64.0FB-IoU: 76.7learnable parameters (million): 2.5
few-shot-semantic-segmentation-on-pascal-5i-1-shotHSNet (VGG-16)#79Mean IoU: 59.7FB-IoU: 73.4
few-shot-semantic-segmentation-on-coco-20i-1-shotHSNet (ResNet-101)#53Mean IoU: 41.2FB-IoU: 69.1learnable parameters (million): 2.5
few-shot-semantic-segmentation-on-coco-20i-1-shotHSNet (ResNet-50)#56Mean IoU: 39.2FB-IoU: 68.2learnable parameters (million): 2.5