Research map / Detection and segmentation / Open-set and few-shot detection
Interclass Prototype Relation for Few-Shot Segmentation
Atsuro Okazawa
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add IPRM (interclass prototype relation loss) (ablation) | IPRNet baseline · COCO-20i 5-shot ResNet-50 | mIoU: 46.7 -> 49.3 | Improved |
| add RCM (respective classifier) alone (ablation) | IPRNet baseline · COCO-20i 5-shot ResNet-50 | mIoU: 46.7 -> 48.6 | Improved |
| add RCM on top of IPRM (ablation) | IPRNet · COCO-20i 5-shot ResNet-50 | mIoU: 49.3 -> 51.1 | Improved |
| add IPRM on top of RCM (ablation) | IPRNet · COCO-20i 5-shot ResNet-50 | mIoU: 48.6 -> 51.1 | Improved |
| use IPRNet instead of HSNet | HSNet · Pascal-5i 1-shot ResNet-50 | mIoU: 64.0 -> 65.7 | Improved |
| use IPRNet instead of HSNet | HSNet · Pascal-5i 5-shot ResNet-50 | mIoU: 69.5 -> 70.2 | Improved |
| use IPRNet instead of HSNet | HSNet · COCO-20i 1-shot ResNet-50 | mIoU: 39.2 -> 45.3 | Improved |
| use IPRNet instead of HSNet | HSNet · COCO-20i 5-shot ResNet-101 | mIoU: 49.5 -> 53.3 | Improved |
| add IPRM and RCM for person class (ablation) | IPRNet baseline · COCO-20i 5-shot per-class IoU, person | IoU: 30.3 -> 51.0 | Improved |
| add IPRM and RCM for kite class (ablation) | IPRNet baseline · COCO-20i 5-shot per-class IoU, kite | IoU: 51.3 -> 42.4 | Hurt |
| add IPRM and RCM for mouse class (ablation) | IPRNet baseline · COCO-20i 5-shot per-class IoU, mouse | IoU: 50.1 -> 41.9 | Hurt |
| add IPRM and RCM across COCO classes (ablation) | IPRNet baseline · COCO-20i 5-shot per-class IoU | IoU: 10 classes up by more than 10%; 10 classes down by over 1% | Mixed |
Closest papers by meaning
- Part-aware Prototype Network for Few-shot Semantic Segmentation (ECCV 2020 · similarity 0.89)
- Cross-Domain Few-Shot Semantic Segmentation (ECCV 2022 · similarity 0.88)
- Intermediate Prototype Mining Transformer for Few-Shot Semantic Segmentation (NeurIPS 2022 · similarity 0.88)
- Unlocking the Potential of Pre-trained Vision Transformers for Few-Shot Semantic Segmentation through Relationship Descriptors (CVPR 2024 · similarity 0.86)
- Deconfound Semantic Shift and Incompleteness in Incremental Few-shot Semantic Segmentation (AAAI 2025 · similarity 0.86)
- A Surprisingly Simple Approach to Generalized Few-Shot Semantic Segmentation (NeurIPS 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
