Research map / Detection and segmentation / Open-set and few-shot detection
Hypercorrelation Squeeze for Few-Shot Segmentation
Juhong Min, Dahyun Kang, Minsu Cho
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 support feature masking (ablation) | HSNet (ResNet101) · PASCAL-5i 1-shot | mIoU: 61.5 -> 66.2 | Improved |
| add support feature masking (ablation) | HSNet (ResNet101) · PASCAL-5i 5-shot | mIoU: 63.7 -> 70.4 | Improved |
| add support feature masking (ablation) | HSNet (ResNet101) · PASCAL-5i 1-shot FB-IoU | FB-IoU: 72.5 -> 77.6 | Improved |
| add support feature masking (ablation) | HSNet (ResNet101) · PASCAL-5i 5-shot FB-IoU | FB-IoU: 73.8 -> 80.6 | Improved |
| switch backbone from ResNet50 to ResNet101 | HSNet · PASCAL-5i 1-shot | mIoU: 64.0 -> 66.2 | Improved |
| switch backbone from ResNet50 to ResNet101 | HSNet · COCO-20i 1-shot | mIoU: 39.2 -> 41.2 | Improved |
| increase support shots from 1 to 5 | HSNet (ResNet50) · COCO-20i | mIoU: 39.2 -> 46.9 | Improved |
| increase support shots from 1 to 5 | HSNet (ResNet50) · PASCAL-5i | mIoU: 64.0 -> 69.5 | Improved |
| compare against PFENet baseline | HSNet vs PFENet (ResNet101) · COCO-20i 1-shot | mIoU: 38.5 -> 41.2 | Improved |
Closest papers by meaning
- Few-Shot Segmentation via Divide-and-Conquer Proxies (IJCV 2023 · similarity 0.85)
- Few-Shot Segmentation Without Meta-Learning: A Good Transductive Inference Is All You Need? (CVPR 2021 · similarity 0.82)
- Learning What Not To Segment: A New Perspective on Few-Shot Segmentation (CVPR 2022 · similarity 0.82)
- Data Augmentation for Low-Level Vision: CutBlur and Mixture-of-Augmentation (IJCV 2024 · similarity 0.81)
- Evidential Robust Feature Learning for Generalized Few-Shot Segmentation (IJCV 2026 · similarity 0.81)
- A Generative Approach at the Instance-Level for Image Segmentation Under Limited Training Data Conditions (Student Abstract) (AAAI 2025 · similarity 0.80)
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.
