Research map / Detection and segmentation / CNN features
Progressive Neighborhood Aggregation for Semantic Segmentation Refinement
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 PNA to PSP | PSP on Pascal VOC 2012 val · Pascal VOC 2012 val | mIoU (%): 78.8% vs. 79.9% | Helped |
| Add PNA to DeeplabV3 | DeeplabV3 with ResNet101 on CityScapes · CityScapes val | mIoU (%): 80.5% vs. 81.3% | Helped |
| Add PNA to FPN | FPN on DeepGlobe · DeepGlobe | mIoU (%): 73.3% vs. 74.0% | Helped |
| Add PNA to FPN | FPN on Trans10k validation · Trans10k val | mIoU (%): 89.5% and 90.7% | Helped |
| Add PNA to FPN | FPN on Trans10k test · Trans10k test | mIoU (%): 90.3% and 91.0% | Helped |
| Add PNA to DeeplabV3 | DeeplabV3 with ResNet50 on COCO-Stuff 10k · COCO-Stuff 10k | mIoU (%): 34.7% and 37.0% | Helped |
| Add SAM after NAM (ablation) | FCN-OS-32 with ResNet50 on Pascal VOC 2012 · Pascal VOC 2012 val | mIoU (%): 76.9 -> 78.6 | Helped |
| Add NAM to FCN-OS-32 (ablation) | FCN-OS-32 with ResNet50 on Pascal VOC 2012 · Pascal VOC 2012 val | mIoU (%): 67.1% vs. 76.9% | Helped |
| Increase NAM window size from 5 to 7 (ablation) | PNA on Pascal VOC 2012 · Pascal VOC 2012 val | mIoU (%): 75.3 -> 75.8 | Helped |
| Increase NAM window size from 7 to 9 (ablation) | PNA on Pascal VOC 2012 · Pascal VOC 2012 val | mIoU (%): 75.8 -> 76.9 | Helped |
| Increase NAM window size from 9 to 11 (ablation) | PNA on Pascal VOC 2012 · Pascal VOC 2012 val | mIoU (%): 76.9 -> 76.9 | No clear effect |
Closest papers by meaning
- Regional Semantic Contrast and Aggregation for Weakly Supervised Semantic Segmentation (CVPR 2022 · similarity 0.87)
- Full-Resolution Residual Networks for Semantic Segmentation in Street Scenes (CVPR 2017 · similarity 0.87)
- Rethinking BiSeNet for Real-Time Semantic Segmentation (CVPR 2021 · similarity 0.87)
- MaSS13K: A Matting-level Semantic Segmentation Benchmark (CVPR 2025 · similarity 0.86)
- Hierarchy-Agnostic Unsupervised Segmentation: Parsing Semantic Image Structure (NeurIPS 2024 · similarity 0.86)
- Learning Hierarchical Image Segmentation For Recognition and By Recognition (ICLR 2024 · similarity 0.86)
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.