Research map / Detection and segmentation / Open-set and few-shot detection
Learning Meta-Class Memory for Few-Shot Semantic Segmentation
Zhonghua Wu, Xiangxi Shi, Guosheng Lin, Jianfei Cai
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 the meta-class memory module instead of direct middle-level features (ablation) | MM-Net · PASCAL-5i 1-shot | mIoU: 58.9 -> 61.8 | Improved |
| add middle-level features and meta-class memory over foreground confidence map alone (ablation) | MM-Net · PASCAL-5i 1-shot | mIoU: 42.8 -> 61.8 | Improved |
| use 50 instead of 20 meta-class embeddings (ablation) | MM-Net · PASCAL-5i 1-shot | mIoU: 61.7 -> 61.8 | No clear effect |
| use 50 instead of 100 meta-class embeddings (ablation) | MM-Net · PASCAL-5i 1-shot | mIoU: 61.0 -> 61.8 | Improved |
| fuse 2nd and 3rd level features for memory instead of 3rd only (ablation) | MM-Net · PASCAL-5i 1-shot | mIoU: 59.6 -> 61.8 | Improved |
| fuse 2+3 level features instead of separate memories (2,3) (ablation) | MM-Net · PASCAL-5i 1-shot | mIoU: 60.0 -> 61.8 | Improved |
| add Activation Propagation Module instead of global average pooling (ablation) | MM-Net · PASCAL-5i 1-shot | mIoU: 61.2 -> 61.8 | Improved |
| add memory reconstruction loss (ablation) | MM-Net · PASCAL-5i 1-shot | mIoU: 60.7 -> 61.8 | Improved |
| apply reconstruction loss on support only instead of support and query (ablation) | MM-Net · PASCAL-5i 1-shot | mIoU: 61.6 -> 61.8 | No clear effect |
| add Quality Measurement Module for k-shot support weighting (ablation) | MM-Net · PASCAL-5i 5-shot | mIoU: 62.6 -> 63.4 | Improved |
| add Foreground Confidence Module (ablation) | MM-Net · PASCAL-5i 1-shot | mIoU: 58.9 -> 61.8 | Improved |
| use QMM instead of CANet-style attention for fusing support maps (ablation) | MM-Net · PASCAL-5i 5-shot | mIoU: 61.7 -> 63.4 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- CRNet — CRNet: Cross-Reference Networks for Few-Shot Segmentation (CVPR 2020) · better on 2 of 2 shared results
- CANet — CANet: Class-Agnostic Segmentation Networks With Iterative Refinement and Attentive Few-Shot Learning (CVPR 2019) · better on 1 of 1 shared result
- PANet — PANet: Few-Shot Image Semantic Segmentation With Prototype Alignment (ICCV 2019) · better on 1 of 1 shared result
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
- Remember the Difference: Cross-Domain Few-Shot Semantic Segmentation via Meta-Memory Transfer (CVPR 2022 · similarity 0.91)
- Simpler Is Better: Few-Shot Semantic Segmentation With Classifier Weight Transformer (ICCV 2021 · similarity 0.91)
- Cross-Domain Few-Shot Semantic Segmentation (ECCV 2022 · similarity 0.90)
- Label-Efficient Few-Shot Semantic Segmentation with Unsupervised Meta-Training (AAAI 2024 · similarity 0.89)
- A Surprisingly Simple Approach to Generalized Few-Shot Semantic Segmentation (NeurIPS 2024 · similarity 0.89)
- Unlocking the Potential of Pre-trained Vision Transformers for Few-Shot Semantic Segmentation through Relationship Descriptors (CVPR 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.
