Research map / Detection and segmentation / Weakly supervised detection and segmentation
SILCO: Show a Few Images, Localize the Common Object
Tao Hu, Pascal Mettes, Jia-Hong Huang, Cees G. M. Snoek
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 |
|---|---|---|---|
| use image-wise instead of global spatial similarity (ablation) | SILCO · CL-VOC-12 | mAP (%): 53.60 -> 55.71 | Improved |
| add spatial similarity module over global average pooling (ablation) | SILCO · CL-VOC-07 | mAP (%): 53.67 -> 55.82 | Improved |
| add spatial similarity module over global average pooling (ablation) | SILCO · CL-VOC-12 | mAP (%): 54.08 -> 55.71 | Improved |
| add spatial similarity module over global average pooling (ablation) | SILCO · CL-COCO | mAP (%): 9.94 -> 13.10 | Improved |
| add feature reweighting on top of spatial similarity (ablation) | SILCO · CL-VOC-07 | mAP (%): 55.82 -> 56.82 | Improved |
| add feature reweighting on top of spatial similarity (ablation) | SILCO · CL-VOC-12 | mAP (%): 55.71 -> 56.86 | Improved |
| add feature reweighting on top of spatial similarity (ablation) | SILCO · CL-COCO | mAP (%): 13.10 -> 13.40 | Improved |
| add SSM and FRM for small objects (ablation) | SILCO vs GAP · CL-VOC-12 small objects | mAP (%): 9.60 -> 10.99 | Improved |
| add SSM and FRM for medium objects (ablation) | SILCO vs GAP · CL-VOC-12 medium objects | mAP (%): 10.87 -> 13.85 | Improved |
| add SSM and FRM for large objects (ablation) | SILCO vs GAP · CL-VOC-12 large objects | mAP (%): 28.64 -> 29.24 | Improved |
| insert a corrupted support image (5-shot) | SILCO · CL-VOC-12 | mAP (%): 56.86 -> 56.53 | No clear effect |
| insert a corrupted support image (5-shot) | SILCO · CL-COCO | mAP (%): 13.40 -> 13.03 | No clear effect |
Closest papers by meaning
- Few-Shot Transformation of Common Actions Into Time and Space (CVPR 2021 · similarity 0.89)
- Localizing the Common Action Among a Few Videos (ECCV 2020 · similarity 0.88)
- Few-Shot Learning With Localization in Realistic Settings (CVPR 2019 · similarity 0.87)
- Meta-Learning to Detect Rare Objects (ICCV 2019 · similarity 0.87)
- StarNet: towards Weakly Supervised Few-Shot Object Detection (AAAI 2021 · similarity 0.87)
- Few-Shot Visual Relationship Co-Localization (ICCV 2021 · similarity 0.87)
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.
