Research map / Detection and segmentation / CNN features
Spatial Pyramid Based Graph Reasoning for Semantic 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 |
|---|---|---|---|
| Add graph convolution without the attention diagonal matrix or Laplacian identity term (ablation) | FCN with ResNet-50 · Cityscapes | mIoU: 76.34 -> 77.98 | Helped |
| Add a learnable data-independent diagonal matrix (ablation) | Simplest GCN · Cityscapes | mIoU: 77.98 -> 78.58; 0.60 | Helped |
| Make the diagonal matrix data-dependent (ablation) | GCN with a data-independent diagonal matrix · Cityscapes | mIoU: 78.58 -> 79.05; 0.47 | Helped |
| Add the identity term to the Laplacian (ablation) | GCN with a data-dependent diagonal matrix · Cityscapes | mIoU: 79.05 -> 79.42 | Helped |
| Add the spatial pyramid (ablation) | GCN with the identity term · Cityscapes | mIoU: 79.42 -> 79.93; 0.51 | Helped |
| Use the full SpyGR method instead of DANet | Semantic segmentation · Cityscapes test | mIoU: 81.5 -> 81.6 | Helped |
| Use the full SpyGR method instead of DANet | Semantic segmentation · COCO Stuff test | mIoU: 39.7 -> 39.9 | Helped |
| Use the full SpyGR method instead of DANet | Semantic segmentation · PASCAL Context test | mIoU: 52.6 -> 52.8 | Helped |
Closest papers by meaning
- Symbolic Graph Reasoning Meets Convolutions (NeurIPS 2018 · similarity 0.88)
- Beyond Grids: Learning Graph Representations for Visual Recognition (NeurIPS 2018 · similarity 0.87)
- Joint Semantic Segmentation and Boundary Detection Using Iterative Pyramid Contexts (CVPR 2020 · similarity 0.87)
- Coupled Segmentation and Edge Learning via Dynamic Graph Propagation (NeurIPS 2021 · similarity 0.86)
- Bridge the Points: Graph-based Few-shot Segment Anything Semantically (NeurIPS 2024 · similarity 0.85)
- Integrating Low-Level Visual Cues for Enhanced Unsupervised Semantic Segmentation (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.