Metric: learnable parameters (million) (higher is better)
| # | Model↕ | learnable parameters (million)▼ | Augmentations | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | DCAMA (ResNet-101) | 47.7 | No | Dense Cross-Query-and-Support Attention Weighted... | 2022-07-18 | Code |
| 2 | PPNet (ResNet-50) | 31.5 | No | Part-aware Prototype Network for Few-shot Semant... | 2020-07-13 | Code |
| 3 | BAM (ResNet-50) | 26.7 | No | Learning What Not to Segment: A New Perspective ... | 2022-03-15 | Code |
| 4 | CyCTR (ResNet-50) | 15.4 | No | Few-Shot Segmentation via Cycle-Consistent Trans... | 2021-06-04 | Code |
| 5 | PFENet (ResNet-101) | 10.3 | No | Prior Guided Feature Enrichment Network for Few-... | 2020-08-04 | Code |
| 6 | HSNet (ResNet-101) | 2.5 | No | Hypercorrelation Squeeze for Few-Shot Segmentation | 2021-04-04 | Code |
| 7 | HSNet (ResNet-50) | 2.5 | No | Hypercorrelation Squeeze for Few-Shot Segmentation | 2021-04-04 | Code |
| 8 | MSDNet (ResNet-101) | 1.5 | No | MSDNet: Multi-Scale Decoder for Few-Shot Semanti... | 2024-09-17 | Code |
| 9 | MSDNet (ResNet-50) | 1.5 | No | MSDNet: Multi-Scale Decoder for Few-Shot Semanti... | 2024-09-17 | Code |