Research map / Detection and segmentation / CNN features
Efficient Piecewise Training of Deep Structured Models 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 sliding pyramid pooling (ablation) | FullyConvNet Baseline · NYUDv2 | IoU: 30.5 -> 32.4 | Helped |
| Add multi-scale input (ablation) | FullyConvNet with sliding pyramid pooling · NYUDv2 | IoU: 32.4 -> 37.0 | Helped |
| Add dense-CRF boundary refinement (ablation) | Multi-scale FullyConvNet with sliding pyramid pooling · NYUDv2 | IoU: 37.0 -> 38.3 | Helped |
| Add CNN contextual pairwise potentials (ablation) | NYUDv2 model with boundary refinement · NYUDv2 | IoU: 38.3 -> 40.6 | Helped |
| Add multi-scale inputs and sliding pyramid pooling to the conventional fully convolutional baseline (ablation) | FullyConvNet Baseline · NYUDv2 | IoU: 30.5 -> 37.0 | Helped |
| Add CNN-based pairwise potentials for patch-patch context (ablation) | Contextual segmentation model · NYUDv2 | IoU: 38.3 -> 40.6 | Helped |
| Add middle-layer refinement features (ablation) | VOC+COCO model · PASCAL VOC 2012 test | IoU: 77.2 -> 78.0 | Helped |
| Train with extra COCO images | ours · PASCAL VOC 2012 test | IoU: 75.3 -> 77.2 | Helped |
Closest papers by meaning
- Semantic Segmentation via Structured Patch Prediction, Context CRF and Guidance CRF (CVPR 2017 · similarity 0.92)
- Semantic Segmentation With Boundary Neural Fields (CVPR 2016 · similarity 0.91)
- Semantic Image Segmentation With Task-Specific Edge Detection Using CNNs and a Discriminatively Trained Domain Transform (CVPR 2016 · similarity 0.89)
- Hierarchically Gated Deep Networks for Semantic Segmentation (CVPR 2016 · similarity 0.89)
- Unsupervised Semantic Segmentation Through Depth-Guided Feature Correlation and Sampling (CVPR 2024 · similarity 0.88)
- Near, far: Patch-ordering enhances vision foundation models' scene understanding (ICLR 2025 · similarity 0.88)
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.