Research map / Detection and segmentation / CNN features
Blended Grammar Network for Human Parsing
Xiaomei Zhang, Yingying Chen, Bingke Zhu, Jinqiao Wang, Ming Tang
Read the paperCheck 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 PCRNN grammar modules to baseline (ablation) | Baseline · PASCAL-Person-Part val | mIoU: 66.61 -> 70.35 | Improved |
| use PCRNN instead of common conv layers with similar parameters (ablation) | Baseline + CCL · PASCAL-Person-Part val | mIoU: 67.04 -> 72.46 | Improved |
| use conspicuous-to-inconspicuous rule order instead of reversed order (ablation) | Baseline + R-PCRNN · PASCAL-Person-Part val | mIoU: 67.50 -> 70.35 | Improved |
| use PCRNN instead of B-PCRNN (ablation) | Baseline + B-PCRNN · PASCAL-Person-Part val | mIoU: 68.1 -> 70.35 | Improved |
| add blended grammar loss (ablation) | Baseline + PCRNN · PASCAL-Person-Part val | mIoU: 70.35 -> 72.46 | Improved |
| add data augmentation (ablation) | BGNet · PASCAL-Person-Part val | mIoU: 72.46 -> 73.13 | Improved |
| add multi-scale testing (ablation) | BGNet · PASCAL-Person-Part val | mIoU: 73.13 -> 74.42 | Improved |
| use blended architecture instead of parallel PGNet (ablation) | PGNet · PASCAL-Person-Part val | mIoU: 71.02 -> 74.42 | Improved |
| add Rule-1 (head-torso) (ablation) | Baseline · PASCAL-Person-Part val | mIoU: 66.61 -> 68.52 | Improved |
| add Rule-2 (upper-arm) (ablation) | BGNet rules 1 · PASCAL-Person-Part val | mIoU: 68.52 -> 70.64 | Improved |
| add Rule-3 (lower-arm) (ablation) | BGNet rules 1-2 · PASCAL-Person-Part val | mIoU: 70.64 -> 71.22 | Improved |
| add Rule-4 (upper-leg) (ablation) | BGNet rules 1-3 · PASCAL-Person-Part val | mIoU: 71.22 -> 71.93 | Improved |
Closest papers by meaning
- Part-Aware Context Network for Human Parsing (CVPR 2020 · similarity 0.89)
- From Simple to Complex Scenes: Learning Robust Feature Representations for Accurate Human Parsing (TPAMI 2024 · similarity 0.89)
- CDGNet: Class Distribution Guided Network for Human Parsing (CVPR 2022 · similarity 0.89)
- Learning Semantic Neural Tree for Human Parsing (ECCV 2020 · similarity 0.88)
- Parsing Based on Parselets: A Unified Deformable Mixture Model for Human Parsing (TPAMI 2016 · similarity 0.87)
- Hierarchical Information Passing Based Noise-Tolerant Hybrid Learning for Semi-Supervised Human Parsing (AAAI 2021 · similarity 0.87)
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.
