Research map / Detection and segmentation / CNN features
Consistent-Separable Feature Representation for Semantic 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 |
|---|---|---|---|
| add CSFLM (ablation) | PSP · Cityscapes validation | mIoU: 78.65 -> 80.03 | Helped |
| add CSFLM (ablation) | Nonlocal · Cityscapes validation | mIoU: 77.64 -> 78.91 | Helped |
| add CSFLM (ablation) | DeepLabv3+ · Cityscapes validation | mIoU: 78.5 -> 79.87 | Helped |
| use 8, 16, 32, or 64 feature-embedding channels (ablation) | CSFLM on DeepLabv3+ · Cityscapes validation | mIoU: 79.57; 79.78; 79.87; 79.79 | No clear effect |
| replace CAC distance supervision with auxiliary category-mask supervision (ablation) | DeepLabv3+ · Cityscapes validation | mIoU: 78.49 -> 78.11 | Hurt |
| use binary cross-entropy in CAC loss (ablation) | CSFLM on DeepLabv3+ · Cityscapes validation | mIoU: 78.49 -> 78.70 | Helped |
| use Lovasz-hinge in CAC loss (ablation) | CSFLM on DeepLabv3+ · Cityscapes validation | mIoU: 78.49 -> 79.87 | Helped |
| add class-aware margins (ablation) | CSFLM on DeepLabv3+ · Cityscapes validation | mIoU: 79.33 -> 79.87 | Helped |
| add AAM (ablation) | CSFLM on DeepLabv3+ · Cityscapes validation | mIoU: 79.87 -> 80.18 | Helped |
| use ResNet101 instead of ResNet50 (ablation) | DeepLabv3+ with CSFLM and AAM · Cityscapes validation | mIoU: 80.18 -> 81.21 | Helped |
| add multi-grid dilation (ablation) | DeepLabv3+ with CSFLM and AAM · Cityscapes validation | mIoU: 81.21 -> 81.46 | Helped |
| add OHEM to the segmentation loss (ablation) | DeepLabv3+ with CSFLM, AAM, and multi-grid · Cityscapes validation | mIoU: 81.46 -> 81.80 | Helped |
Closest papers by meaning
- Rethinking Semantic Segmentation: A Prototype View (CVPR 2022 · similarity 0.88)
- Renovate Yourself: Calibrating Feature Representation of Misclassified Pixels for Semantic Segmentation (AAAI 2022 · similarity 0.88)
- Perturbed and Strict Mean Teachers for Semi-Supervised Semantic Segmentation (CVPR 2022 · similarity 0.87)
- Towards the Uncharted: Density-Descending Feature Perturbation for Semi-supervised Semantic Segmentation (CVPR 2024 · similarity 0.87)
- Complex Wavelet Mutual Information Loss: A Multi-Scale Loss Function for Semantic Segmentation (ICML 2025 · similarity 0.87)
- Improving Semi-Supervised Semantic Segmentation with Sliced-Wasserstein Feature Alignment and Uniformity (CVPR 2025 · 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.