Research map / Detection and segmentation / CNN features
CascadePSP: Toward Class-Agnostic and Very High-Resolution Segmentation via Global and Local Refinement
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 |
|---|---|---|---|
| refine PSPNet predictions with CascadePSP | PSPNet · PASCAL VOC 2012 val | IoU (%): 90.92 -> 92.86 | Helped |
| refine PSPNet predictions with CascadePSP | PSPNet · BIG test | mBA (%): 59.63 -> 75.32 | Helped |
| refine PSPNet predictions with CascadePSP | PSPNet · ADE20K val | mIoU (%): 43.10 -> 43.83 | Helped |
| add the OS4 intermediate output to OS8 and OS1 (ablation) | CascadePSP refinement module · PASCAL VOC 2012 | IoU (%): 88.85 -> 89.01 | Helped |
| add the OS4 intermediate output to OS8 and OS1 (ablation) | CascadePSP refinement module · PASCAL VOC 2012 | mBA (%): 71.88 -> 72.10 | Helped |
| add the stride-1 gradient loss (ablation) | CascadePSP refinement module · PASCAL VOC 2012 | mBA (%): 71.36 -> 72.10 | Helped |
| add the stride-1 gradient loss (ablation) | CascadePSP refinement module · PASCAL VOC 2012 | IoU (%): 88.84 -> 89.01 | Helped |
| use both Global and Local refinement steps (ablation) | CascadePSP · BIG val | IoU (%): 91.86 -> 92.01 | Helped |
| use both Global and Local refinement steps (ablation) | CascadePSP · BIG val | mBA (%): 73.10 -> 75.59 | Helped |
| remove the Global step (ablation) | CascadePSP · BIG val | IoU (%): 92.01 -> 91.35 | Hurt |
| remove the Local step (ablation) | CascadePSP · BIG val | mBA (%): 75.59 -> 73.10 | Hurt |
| use Global and Local steps instead of Global step alone (ablation) | CascadePSP · BIG val | mBA (%): 73.10 -> 75.59 | Helped |
Closest papers by meaning
- Patch Proposal Network for Fast Semantic Segmentation of High-Resolution Images (AAAI 2020 · similarity 0.87)
- Hierarchy-Agnostic Unsupervised Segmentation: Parsing Semantic Image Structure (NeurIPS 2024 · similarity 0.86)
- MRFP: Learning Generalizable Semantic Segmentation from Sim-2-Real with Multi-Resolution Feature Perturbation (CVPR 2024 · similarity 0.85)
- Cascade-CLIP: Cascaded Vision-Language Embeddings Alignment for Zero-Shot Semantic Segmentation (ICML 2024 · similarity 0.85)
- Unsupervised Semantic Segmentation Through Depth-Guided Feature Correlation and Sampling (CVPR 2024 · similarity 0.85)
- Stronger Fewer & Superior: Harnessing Vision Foundation Models for Domain Generalized Semantic Segmentation (CVPR 2024 · 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.