Research map / Detection and segmentation / Open-set and few-shot detection
Exemplar-Based Open-Set Panoptic Segmentation Network
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 exemplar clustering and progressive exemplar mining | Panoptic FPN with Void-train · COCO val, K=5% | Unknown PQ: 10.0 -> 23.1 | Helped |
| Add exemplar clustering and progressive exemplar mining | Panoptic FPN with Void-train · COCO val, K=10% | Unknown PQ: 8.5 -> 17.9 | Helped |
| Add exemplar clustering and progressive exemplar mining | Panoptic FPN with Void-train · COCO val, K=20% | Unknown PQ: 7.8 -> 11.3 | Helped |
| Train void-region boxes as a new label rather than ignore them (ablation) | Panoptic FPN · COCO val, K=20% | Unknown PQ: 3.7 -> 7.8 | Helped |
| Apply negative known-class supervision to void-region boxes (ablation) | Panoptic FPN · COCO val, K=20% | Unknown PQ: 4.0 -> 7.2 | Helped |
| Use 128 clusters rather than 64 (ablation) | EOPSN · COCO val, K=10% | Unknown PQ: 12.9 -> 17.9 | Helped |
| Use 128 clusters rather than 256 (ablation) | EOPSN · COCO val, K=10% | Unknown PQ: 12.6 -> 17.9 | Helped |
| Cluster every 200 iterations rather than every 100 (ablation) | EOPSN · COCO val, K=10% | Unknown PQ: 8.2 -> 17.9 | Helped |
| Cluster every 200 iterations rather than every 400 (ablation) | EOPSN · COCO val, K=10% | Unknown PQ: 14.6 -> 17.9 | Helped |
| Use large and medium proposals together rather than large proposals only (ablation) | EOPSN · COCO val, K=10% | Unknown PQ: 13.5 -> 17.9 | Helped |
| Use large and medium proposals together rather than medium proposals only (ablation) | EOPSN · COCO val, K=10% | Unknown PQ: 12.5 -> 17.9 | Helped |
| Use large and medium proposals rather than small proposals only (ablation) | EOPSN · COCO val, K=10% | Unknown PQ: 0.3 -> 17.9 | Helped |
Closest papers by meaning
- Revisiting Open-Set Panoptic Segmentation (AAAI 2024 · similarity 0.91)
- Auto-Panoptic: Cooperative Multi-Component Architecture Search for Panoptic Segmentation (NeurIPS 2020 · similarity 0.88)
- An End-To-End Network for Panoptic Segmentation (CVPR 2019 · similarity 0.88)
- EOV-Seg: Efficient Open-Vocabulary Panoptic Segmentation (AAAI 2025 · similarity 0.88)
- Panoptic Feature Pyramid Networks (CVPR 2019 · similarity 0.88)
- Unifying Training and Inference for Panoptic Segmentation (CVPR 2020 · 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.