Image Segmentation on COCO-20i (val)
77.7FB-IoU (%)HDMNet
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| HDMNetBackbone=ResNet-50, Shot Count=5-shot, #learnable params=4.2M2023.03 | 77.7 | — | |
| BAMBackbone=ResNet-50, Shot Count=5-shot, #learnable params=4.9M2023.03 | 73.3 | — | |
| HDMNetBackbone=ResNet-50, Shot Count=1-shot, #learnable params=4.2M2023.03 | 72.2 | — | |
| BAMBackbone=ResNet-50, Shot Count=1-shot, #learnable params=4.9M2023.03 | 71.1 | — | |
| HSNetBackbone=ResNet-50, Shot Count=5-shot, #learnable params=2.6M2023.03 | 70.7 | — | |
| HSNetBackbone=ResNet-50, Shot Count=1-shot, #learnable params=2.6M2023.03 | 68.2 | — | |
| ASGNetBackbone=ResNet-50, Shot Count=5-shot, #learnable params=10.4M2023.03 | 67 | — | |
| ASGNetBackbone=ResNet50, Shots=5-shot2021.04 | 66.96 | 42.48 | |
| DANBackbone=ResNet101, Shots=5-shot2021.04 | 63.9 | 29.6 | |
| DANBackbone=ResNet101, Shots=1-shot2021.04 | 62.3 | 24.2 | |
| PFENetBackbone=ResNet101, Shots=5-shot2021.04 | 61.9 | 37.4 | |
| ASGNetBackbone=ResNet-50, Shot Count=1-shot, #learnable params=10.4M2023.03 | 60.4 | — | |
| ASGNetBackbone=ResNet50, Shots=1-shot2021.04 | 60.39 | 34.56 | |
| PFENetBackbone=ResNet101, Shots=1-shot2021.04 | 58.6 | 32.4 | |
| FWBBackbone=ResNet101, Shots=1-shot2021.04 | — | 21.19 | |
| FWBBackbone=ResNet101, Shots=5-shot2021.04 | — | 23.65 | |
| RPMMsBackbone=ResNet50, Shots=1-shot2021.04 | — | 30.58 | |
| RPMMsBackbone=ResNet50, Shots=5-shot2021.04 | — | 35.52 |