Research map / Detection and segmentation / CNN features
Patch Proposal Network for Fast Semantic Segmentation of High-Resolution Images
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 |
|---|---|---|---|
| add PPN patch selection to the enhance (all patches refined equally) model (ablation) | GRNet · CRAG | mIoU (%): 87.7 -> 88.9 | Helped |
| add refinement branch with PPN-selected patches to global-only network (ablation) | GlobalNet · CRAG | mIoU (%): 88.5 -> 88.9 | Helped |
| add global branch to local-only network (ablation) | LocalNet · CRAG | mIoU (%): 75.5 -> 88.5 | Helped |
| refine all patches equally without PPN, added to global network (ablation) | GlobalNet · CRAG | mIoU (%): 88.5 -> 87.7 | Hurt |
| increase number of patches from 4 to 16 (ablation) | GRNet · DeepGlobe | mIoU (%): 71.5 -> 71.9 | Helped |
| increase number of patches from 16 to 64 (ablation) | GRNet · DeepGlobe | mIoU (%): 71.9 -> 72.4 | Helped |
| replace GLNet-style equal patch fusion with PPN-guided GRNet | GLNet · DeepGlobe | FPS: 0.5 -> 12.9 | Helped |
| replace GLNet with GRNet | GLNet · DeepGlobe | mIoU (%): 71.6 -> 71.9 | Helped |
| embed PPN with refinement into BiSeNet | BiSeNet · Cityscapes val | mIoU (%): 74.7 -> 75.2 | Helped |
Closest papers by meaning
- Collaborative Global-Local Networks for Memory-Efficient Segmentation of Ultra-High Resolution Images (CVPR 2019 · similarity 0.89)
- ISDNet: Integrating Shallow and Deep Networks for Efficient Ultra-High Resolution Segmentation (CVPR 2022 · similarity 0.88)
- Toward Real Ultra Image Segmentation: Leveraging Surrounding Context to Cultivate General Segmentation Model (NeurIPS 2024 · similarity 0.87)
- CascadePSP: Toward Class-Agnostic and Very High-Resolution Segmentation via Global and Local Refinement (CVPR 2020 · similarity 0.87)
- ALGM: Adaptive Local-then-Global Token Merging for Efficient Semantic Segmentation with Plain Vision Transformers (CVPR 2024 · similarity 0.86)
- Ultra-High Resolution Segmentation via Boundary-Enhanced Patch-Merging Transformer (AAAI 2025 · similarity 0.85)
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.