Research map / Detection and segmentation / CNN features
Learning Semantic Neural Tree for Human Parsing
Ruyi Ji, Dawei Du, Libo Zhang, Longyin Wen, Yanjun Wu, Chen Zhao, Feiyue Huang, Siwei Lyu
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 the prediction module for the final parsing output (ablation) | SNT · LIP val | mIoU: 50.02 -> 54.73 | Improved |
| use deformable convolutions in the prediction module (ablation) | SNT · LIP val | mIoU: 54.42 -> 54.73 | Improved |
| add skip connections from backbone residual blocks to semantic aggregation (ablation) | SNT · LIP val | mIoU: 53.32 -> 54.73 | Improved |
| add attention mask in attention routing module (ablation) | SNT w/o skip · LIP val | mIoU: 52.15 -> 53.32 | Improved |
| increase tree height from 0 to 3 (ablation) | SNT · LIP val | mIoU: 46.34 -> 54.73 | Improved |
| increase tree height from 3 to 4 (ablation) | SNT · LIP val | mIoU: 54.73 -> 51.42 | Hurt |
| increase tree height from 2 to 3 (ablation) | SNT · LIP val | mIoU: 53.32 -> 54.73 | Improved |
| use SNT instead of CE2P (w flip) | human parsing · LIP val | mIoU: 53.10 -> 54.73 | Improved |
| use SNT instead of DPC | human parsing · Pascal-Person-Part val | mIoU: 71.34 -> 71.59 | Improved |
| use SNT instead of M-CE2P at strict IoU 0.7 | multiple human parsing · CIHP val | AP^r_0.7: 29.74 -> 33.00 | Improved |
| use SNT instead of M-CE2P | multiple human parsing · MHP-v2 val | PCP_0.5: 43.77 -> 43.50 | No clear effect |
Closest papers by meaning
- Semantic Human Parsing via Scalable Semantic Transfer Over Multiple Label Domains (CVPR 2023 · similarity 0.90)
- Hierarchical Human Semantic Parsing with Comprehensive Part-Relation Modeling (TPAMI 2021 · similarity 0.89)
- Part-Aware Context Network for Human Parsing (CVPR 2020 · similarity 0.89)
- Hierarchical Human Parsing With Typed Part-Relation Reasoning (CVPR 2020 · similarity 0.88)
- Learning Compositional Neural Information Fusion for Human Parsing (ICCV 2019 · similarity 0.88)
- CDGNet: Class Distribution Guided Network for Human Parsing (CVPR 2022 · 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.
