Research map / Detection and segmentation / Open-set and few-shot detection
Panoptic Segmentation
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 |
|---|---|---|---|
| Merge thing and stuff predictions, resolving overlaps in favor of things | Independent segmentation outputs · Cityscapes | PQ St: 66.6 -> 66.4 | Hurt |
| Merge thing and stuff predictions, resolving overlaps in favor of things | Independent segmentation outputs · ADE20k | PQ St: 27.4 -> 24.5 | Hurt |
| Merge thing and stuff predictions, resolving overlaps in favor of things | Independent segmentation outputs · Mapillary Vistas | PQ St: 43.7 -> 41.8 | Hurt |
| Resolve overlaps between predicted instance segments to make them non-overlapping | Instance segmentation predictions · Cityscapes | AP: 36.4 -> 33.1 | Hurt |
| Resolve overlaps between predicted instance segments to make them non-overlapping | Instance segmentation predictions · ADE20k | AP: 30.1 -> 24.8 | Hurt |
| Use a matching threshold of 0.25 instead of 0.5 (ablation) | Panoptic quality evaluation · Cityscapes, ADE20k, and Vistas human annotations | PQ | No clear effect |
| Lower the recognition-error penalty alpha below 0.5 (ablation) | Panoptic quality evaluation · Human consistency study | RQ | Helped |
| Use a lower alpha than the default 0.5 (ablation) | Panoptic quality evaluation · Human consistency study | SQ | No clear effect |
| Evaluate segmentation by object scale rather than pooling sizes | Human panoptic segmentation · Cityscapes, ADE20k, and Vistas | PQ: Cityscapes: 35.1 / 62.3 / 84.8; ADE20k: 49.9 / 69.4 / 79.0; Vistas: 35.6 / 47.7 / 69.4 | Mixed |
Closest papers by meaning
- Part-Aware Panoptic Segmentation (CVPR 2021 · similarity 0.91)
- Real-Time Panoptic Segmentation From Dense Detections (CVPR 2020 · similarity 0.91)
- Video Panoptic Segmentation (CVPR 2020 · similarity 0.90)
- Scene-Centric Unsupervised Panoptic Segmentation (CVPR 2025 · similarity 0.90)
- Task-aligned Part-aware Panoptic Segmentation through Joint Object-Part Representations (CVPR 2024 · similarity 0.88)
- Fully Data-Driven Pseudo Label Estimation for Pointly-Supervised Panoptic Segmentation (AAAI 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.