Research map / Detection and segmentation / Open-set and few-shot detection
Generalized Few-Shot Semantic Segmentation
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 CAPL to PSPNet | PSPNet, ResNet-50 · Pascal-5^i, GFS-Seg, 1-shot | Total mIoU: 49.54 -> 54.38 | Helped |
| Add CAPL to PSPNet | PSPNet, ResNet-50 · COCO-20^i, GFS-Seg, 1-shot | Total mIoU: 29.06 -> 35.46 | Helped |
| Use cosine similarity for DQCE weighting (ablation) | CAPL on PSPNet · Pascal-5^i, GFS-Seg, 1-shot | Total mIoU: 51.82 -> 52.55 | Helped |
| Switch the DQCE weighting between classifier and query prototypes (ablation) | CAPL on PSPNet · Pascal-5^i, GFS-Seg, 1-shot | Total mIoU: 52.55 -> 48.20 | Hurt |
| Use MLP rather than cosine similarity for SCE weighting (ablation) | CAPL on PSPNet · Pascal-5^i, GFS-Seg, 1-shot | Total mIoU: 49.62 -> 51.63 | Helped |
| Switch the SCE weighting between classifier and support prototypes (ablation) | CAPL on PSPNet · Pascal-5^i, GFS-Seg, 1-shot | Total mIoU: 51.63 -> 50.32 | Hurt |
| Use CAPL training, registration, and evaluation instead of training alignment alone (ablation) | CAPL on PSPNet · Pascal-5^i, GFS-Seg, 1-shot | Total mIoU: 49.65 -> 54.38 | Helped |
| Add CAPL training alignment to registration and evaluation enrichment (ablation) | CAPL on PSPNet · Pascal-5^i, GFS-Seg, 1-shot | Total mIoU: 48.06 -> 54.38 | Helped |
| Add CAPL to PANet | PANet, ResNet-50 · Pascal-5^i, FS-Seg, 5-shot | class mIoU: 55.7 -> 66.1 | Helped |
| Add CAPL to PFENet | PFENet, ResNet-50 · COCO-20^i, FS-Seg, 5-shot | class mIoU: 37.5 -> 48.3 | Helped |
Closest papers by meaning
- A Surprisingly Simple Approach to Generalized Few-Shot Semantic Segmentation (NeurIPS 2024 · similarity 0.95)
- Enhancing Generalized Few-Shot Semantic Segmentation via Effective Knowledge Transfer (AAAI 2025 · similarity 0.93)
- Label-Efficient Few-Shot Semantic Segmentation with Unsupervised Meta-Training (AAAI 2024 · similarity 0.92)
- Unlocking the Potential of Pre-trained Vision Transformers for Few-Shot Semantic Segmentation through Relationship Descriptors (CVPR 2024 · similarity 0.91)
- APSeg: Auto-Prompt Network for Cross-Domain Few-Shot Semantic Segmentation (CVPR 2024 · similarity 0.90)
- Bridge the Points: Graph-based Few-shot Segment Anything Semantically (NeurIPS 2024 · similarity 0.89)
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.