Research map / Detection and segmentation / CNN features
Structural and Statistical Texture Knowledge Distillation for Semantic Segmentation
Deyi Ji, Haoran Wang, Mingyuan Tao, Jianqiang Huang, Xian-Sheng Hua, Hongtao Lu
Check your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 structural and statistical texture distillation | PSPNet-R18 · Cityscapes val | mIoU: 69.10 -> 75.15 | Improved |
| add structural texture knowledge (ablation) | PSPNet-R18 with response knowledge · Cityscapes val | mIoU: 72.47 -> 74.10 | Improved |
| add statistical texture knowledge (ablation) | PSPNet-R18 with response knowledge · Cityscapes val | mIoU: 72.47 -> 74.69 | Improved |
| combine structural and statistical texture knowledge (ablation) | PSPNet-R18 with response knowledge · Cityscapes val | mIoU: 72.47 -> 75.15 | Improved |
| increase contourlet decomposition from 1 to 2 levels (ablation) | CDM in PSPNet-R18 · Cityscapes val | mIoU: 73.44 -> 74.10 | Improved |
| increase contourlet decomposition from 2 to 3 levels (ablation) | CDM in PSPNet-R18 · Cityscapes val | mIoU: 74.10 -> 74.11 | No clear effect |
| use anchor-based adaptive importance sampling instead of global extraction (ablation) | DTIEM in PSPNet-R18 · Cityscapes val | mIoU: 73.57 -> 74.06 | Improved |
| add heuristic iterative quantization-level initialization (ablation) | DTIEM with adaptive sampling · Cityscapes val | mIoU: 74.06 -> 74.25 | Improved |
| add denoised intensity-limited quantization (ablation) | DTIEM with adaptive sampling and heuristic initialization · Cityscapes val | mIoU: 74.25 -> 74.69 | Improved |
| add SSTKD | DeepLab-R18 · Cityscapes val | mIoU: 73.37 -> 76.13 | Improved |
| add SSTKD | PSPNet-R18 · Pascal VOC 2012 val | mIoU: 65.42 -> 70.98 | Improved |
| add SSTKD | PSPNet-R18 · ADE20K val | mIoU: 24.65 -> 29.19 | Improved |
Closest papers by meaning
- Structural and Statistical Texture Knowledge Distillation and Learning for Segmentation (TPAMI 2025 · similarity 0.97)
- Learning Statistical Texture for Semantic Segmentation (CVPR 2021 · similarity 0.89)
- Structured Knowledge Distillation for Semantic Segmentation (CVPR 2019 · similarity 0.88)
- Distilling Knowledge from Heterogeneous Architectures for Semantic Segmentation (AAAI 2025 · similarity 0.87)
- Label-Guided Knowledge Distillation for Continual Semantic Segmentation on 2D Images and 3D Point Clouds (ICCV 2023 · similarity 0.87)
- Integrating Low-Level Visual Cues for Enhanced Unsupervised Semantic Segmentation (AAAI 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.
