Research map / Detection and segmentation / CNN features
Auto-Panoptic: Cooperative Multi-Component Architecture Search for Panoptic Segmentation
Read the paperCheck 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 |
|---|---|---|---|
| Search all major components jointly, including backbone, heads, and inter-branch fusion (ablation) | Auto-Panoptic · COCO val | PQ: 43.9 -> 44.8 | Helped |
| Add a searched backbone to the searched head and inter-branch fusion (ablation) | Cooperative multi-component search · COCO val | PQ: 43.3 -> 44.8 | Helped |
| Add inter-branch fusion to the searched backbone and head (ablation) | Cooperative multi-component search · COCO val | PQ: 43.9 -> 44.8 | Helped |
| Use diverse searched convolution operators instead of plain 3x3 convolutions in the head (ablation) | Auto-Panoptic · COCO val | PQ: 42.9 -> 44.8 | Helped |
| Use the jointly searched architecture instead of the best of five randomly selected architectures (ablation) | Auto-Panoptic · COCO val | PQ: 41.1 -> 44.8 | Helped |
| Use Path-Priority search instead of evolutionary search | Architecture search · COCO val | PQ: 43.8 -> 44.8 | Helped |
| Use Path-Priority search instead of evolutionary search | Architecture search · COCO val | evaluated models: 1000 -> 60 | Helped |
| Use Path-Priority search instead of random search | Architecture search · COCO val | PQ: 42.8 -> 44.8 | Helped |
| Transfer the ADE20K-searched architecture to COCO instead of searching on COCO | Auto-Panoptic · COCO val | PQ: 43.9 vs 44.8 | Hurt |
| Transfer the COCO-searched architecture to ADE20K instead of searching on ADE20K | Auto-Panoptic · ADE20K val | PQ: 32.0 vs 32.4 | Hurt |
| Apply deformable convolution in the semantic segmentation branch | Panoptic-FPN · COCO val | PQ: 39.0 -> 39.9 | Helped |
Closest papers by meaning
- Real-Time Panoptic Segmentation From Dense Detections (CVPR 2020 · similarity 0.91)
- Unifying Training and Inference for Panoptic Segmentation (CVPR 2020 · similarity 0.91)
- Bidirectional Graph Reasoning Network for Panoptic Segmentation (CVPR 2020 · similarity 0.90)
- Attention-Guided Unified Network for Panoptic Segmentation (CVPR 2019 · similarity 0.90)
- Panoptic, Instance and Semantic Relations: A Relational Context Encoder To Enhance Panoptic Segmentation (CVPR 2022 · similarity 0.90)
- LPSNet: A Lightweight Solution for Fast Panoptic Segmentation (CVPR 2021 · similarity 0.90)
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.