Research map / Detection and segmentation / Open-set and few-shot detection
Object-level Correlation for Few-Shot Segmentation
Chunlin Wen, Yu Zhang, Jie Fan, Hongyuan Zhu, Xiu-Shen Wei, Yijun Wang, Zhiqiang Kou, Shuzhou Sun
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 GOMM (general object mining module) (ablation) | baseline · PASCAL-5i 1-shot (ablation, ResNet-50) | mIoU: 67.3 -> 69.1 | Improved |
| add CCM (correlation construction module) (ablation) | baseline · PASCAL-5i 1-shot (ablation, ResNet-50) | mIoU: 67.3 -> 69.9 | Improved |
| add both GOMM and CCM (ablation) | baseline · PASCAL-5i 1-shot (ablation, ResNet-50) | mIoU: 67.3 -> 71.4 | Improved |
| add CCM on top of GOMM (ablation) | GOMM · PASCAL-5i 1-shot (ablation, ResNet-50) | mIoU: 69.1 -> 71.4 | Improved |
| add CAM to mine general objects (ablation) | CCM only · PASCAL-5i 1-shot (ablation, ResNet-50) | mIoU: 69.9 -> 70.3 | Improved |
| add similarity mask to CAM (ablation) | CCM + CAM · PASCAL-5i 1-shot (ablation, ResNet-50) | mIoU: 70.3 -> 70.7 | Improved |
| add cross-attention to CAM (ablation) | CCM + CAM · PASCAL-5i 1-shot (ablation, ResNet-50) | mIoU: 70.3 -> 70.9 | Improved |
| add mask and cross-attention together to CAM (ablation) | CCM + CAM · PASCAL-5i 1-shot (ablation, ResNet-50) | mIoU: 70.3 -> 71.4 | Improved |
| use foreground+background supervised allocation instead of mean (ablation) | CCM · PASCAL-5i 1-shot (ablation, ResNet-50) | mIoU: 70.3 -> 71.4 | Improved |
| use foreground+background allocation instead of cosine similarity (ablation) | CCM · PASCAL-5i 1-shot (ablation, ResNet-50) | mIoU: 70.6 -> 71.4 | Improved |
| add background prototypes to foreground allocation (ablation) | CCM · PASCAL-5i 1-shot (ablation, ResNet-50) | mIoU: 70.5 -> 71.4 | Improved |
Closest papers by meaning
- Cross-Domain Few-Shot Semantic Segmentation (ECCV 2022 · similarity 0.90)
- Hierarchical Dense Correlation Distillation for Few-Shot Segmentation (CVPR 2023 · similarity 0.90)
- Rethinking Few-shot 3D Point Cloud Semantic Segmentation (CVPR 2024 · similarity 0.88)
- A Surprisingly Simple Approach to Generalized Few-Shot Semantic Segmentation (NeurIPS 2024 · similarity 0.88)
- Label-Efficient Few-Shot Semantic Segmentation with Unsupervised Meta-Training (AAAI 2024 · similarity 0.88)
- Unlocking the Potential of Pre-trained Vision Transformers for Few-Shot Semantic Segmentation through Relationship Descriptors (CVPR 2024 · similarity 0.88)
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.
