Research map / Detection and segmentation / Open-set and few-shot detection
Scale-Aware Graph Neural Network for Few-Shot Semantic Segmentation
Guo-Sen Xie, Jie Liu, Huan Xiong, Ling Shao
Check 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 cross-scale graph reasoning over multi-scale query nodes (ablation) | SAGNN · PASCAL-5i | mean-IoU, 1-shot: 59.9 -> 62.1 | Improved |
| Add cross-scale graph reasoning over multi-scale query nodes (ablation) | SAGNN with ResNet-101 · COCO-20i | mean-IoU, 1-shot: 34.1 -> 37.2 | Improved |
| Add self-node collaboration to opposite-node message passing (ablation) | SAGNN · PASCAL-5i | mean-IoU, 1-shot: 61.1 -> 62.1 | Improved |
| Add self-node collaboration to opposite-node message passing (ablation) | SAGNN · PASCAL-5i | mean-IoU, 5-shot: 62.4 -> 62.8 | Improved |
| Use 3 graph scales instead of 2 or 4 (ablation) | SAGNN · PASCAL-5i | mean-IoU, 1-shot: 61.0; 62.1; 61.0 | Mixed |
| Use 2 message-passing steps instead of 1 (ablation) | SAGNN · PASCAL-5i | mean-IoU: 62.1 -> 62.4; 62.8 -> 62.9 | Improved |
| Use 3 message-passing steps instead of 1 (ablation) | SAGNN · PASCAL-5i | mean-IoU: 62.1 -> 62.1; 62.8 -> 62.5 | Mixed |
| Use ResNet-50 rather than VGG-16 or ResNet-101 (ablation) | SAGNN · PASCAL-5i | mean-IoU: 58.4; 62.1; 60.8 (1-shot); 59.3; 62.8; 61.5 (5-shot) | Mixed |
| Average support image features for 5-shot fusion rather than averaging masks or OR-ing masks (ablation) | SAGNN · PASCAL-5i | mean-IoU, 5-shot: 62.5; 61.8 -> 62.8 | Improved |
Closest papers by meaning
- Few-Shot Semantic Segmentation With Cyclic Memory Network (ICCV 2021 · similarity 0.89)
- Part-aware Prototype Network for Few-shot Semantic Segmentation (ECCV 2020 · similarity 0.89)
- Bridge the Points: Graph-based Few-shot Segment Anything Semantically (NeurIPS 2024 · similarity 0.88)
- SAM-Aware Graph Prompt Reasoning Network for Cross-Domain Few-Shot Segmentation (AAAI 2025 · similarity 0.87)
- Label-Efficient Few-Shot Semantic Segmentation with Unsupervised Meta-Training (AAAI 2024 · similarity 0.86)
- Unlocking the Potential of Pre-trained Vision Transformers for Few-Shot Semantic Segmentation through Relationship Descriptors (CVPR 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.
