Research map / Detection and segmentation / Open-set and few-shot detection
Part-aware Prototype Network for Few-shot Semantic Segmentation
Yongfei Liu, Xiangyi Zhang, Songyang Zhang, Xuming He
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 |
|---|---|---|---|
| decompose holistic prototype into part-aware prototypes (without global context) (ablation) | PANet* baseline · COCO-20i split-A 1-way 1-shot | mean-IoU: 22.95 -> 23.35 | Improved |
| add global context to part-aware prototypes (ablation) | PAP · COCO-20i split-A 1-way 1-shot | mean-IoU: 23.35 -> 25.02 | Improved |
| add semantic branch auxiliary loss (ablation) | PAP · COCO-20i split-A 1-way 1-shot | mean-IoU: 25.02 -> 25.71 | Improved |
| add unlabeled data with graph attention refinement (step-2 GNN) (ablation) | PAP + SEM · COCO-20i split-A 1-way 1-shot | mean-IoU: 26.34 -> 27.16 | Improved |
| add unlabeled data without step-2 GNN (ablation) | PAP + SEM · COCO-20i split-A 1-way 1-shot | mean-IoU: 25.71 -> 26.34 | Improved |
| add unlabeled support data (ablation) | PPNet ResNet-50 · PASCAL-5i 1-way 1-shot | mean-IoU: 51.50 -> 52.84 | Improved |
| add unlabeled support data (ablation) | PPNet ResNet-50 · PASCAL-5i 1-way 5-shot | mean-IoU: 61.96 -> 62.97 | Improved |
| add unlabeled support data (ablation) | PPNet ResNet-50 · PASCAL-5i 2-way 1-shot | mean-IoU: 50.24 -> 51.65 | Improved |
| add unlabeled support data (ablation) | PPNet ResNet-50 · COCO-20i split-A 1-way 5-shot | mean-IoU: 36.24 -> 36.73 | Improved |
| replace holistic prototype (PANet*) with PPNet w/o unlabeled data | PANet* ResNet-50 · PASCAL-5i 1-way 5-shot | mean-IoU: 59.26 -> 61.96 | Improved |
| replace PPNet with ResNet-101 backbone instead of ResNet-50 | PPNet · PASCAL-5i 1-way 5-shot | mean-IoU: 62.97 -> 65.10 | Improved |
| PPNet versus multi-scale PGNet in the 1-shot setting | PPNet ResNet-50 · PASCAL-5i 1-way 1-shot | mean-IoU: 52.84 vs 56.00 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PANet* — PANet: Few-Shot Image Semantic Segmentation With Prototype Alignment (ICCV 2019) · better on 9 of 9 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Learning from the Target: Dual Prototype Network for Few Shot Semantic Segmentation (AAAI 2022 · similarity 0.91)
- Cross-Domain Few-Shot Semantic Segmentation (ECCV 2022 · similarity 0.91)
- Generalized Few-Shot Semantic Segmentation (CVPR 2022 · similarity 0.91)
- Unlocking the Potential of Pre-trained Vision Transformers for Few-Shot Semantic Segmentation through Relationship Descriptors (CVPR 2024 · similarity 0.91)
- Prototype Mixture Models for Few-shot Semantic Segmentation (ECCV 2020 · similarity 0.90)
- A Surprisingly Simple Approach to Generalized Few-Shot Semantic Segmentation (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.
