Research map / Detection and segmentation / CNN features
Renovating Parsing R-CNN for Accurate Multiple Human Parsing
Lu Yang, Qing Song, Zhihui Wang, Mengjie Hu, Chun Liu, Xueshi Xin, Wenhe Jia, Songcen Xu
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 GSE-FPN (global semantic enhanced FPN) (ablation) | Parsing R-CNN baseline · CIHP val | mIoU: 56.2 -> 58.2 | Improved |
| add GSE-FPN (ablation) | Parsing R-CNN baseline · CIHP val | AP^p_50: 64.6 -> 67.4 | Improved |
| add PRSN (parsing re-scoring network) (ablation) | Parsing R-CNN baseline · CIHP val | AP^p_50: 64.6 -> 70.3 | Improved |
| add PRSN (ablation) | Parsing R-CNN baseline · CIHP val | mIoU: 56.2 -> 56.4 | No clear effect |
| add both GSE-FPN and PRSN (ablation) | RP R-CNN · CIHP val | AP^p_50: 64.6 -> 71.6 | Improved |
| add GSE-FPN (ablation) | Parsing R-CNN baseline · MHP-v2 val | mIoU: 35.5 -> 37.3 | Improved |
| add PRSN (ablation) | Parsing R-CNN baseline · MHP-v2 val | AP^p_50: 26.6 -> 39.6 | Improved |
| add both GSE-FPN and PRSN (ablation) | RP R-CNN · MHP-v2 val | AP^p_50: 26.6 -> 40.5 | Improved |
| combine semseg and parsing outputs at inference instead of parsing alone (ablation) | RP R-CNN · CIHP val | mIoU: 57.4 -> 58.2 | Improved |
| use ground-truth parsing maps | Parsing R-CNN · CIHP val | AP^p_50: 64.6 -> 91.4 | Improved |
| use ground-truth boxes | Parsing R-CNN · CIHP val | AP^p_50: 64.6 -> 58.0 | Hurt |
| use ground-truth IoU as instance score | Parsing R-CNN · CIHP val | AP^p_50: 64.6 -> 73.7 | Improved |
Closest papers by meaning
- Parsing R-CNN for Instance-Level Human Analysis (CVPR 2019 · similarity 0.90)
- Deep Learning Technique for Human Parsing: A Survey and Outlook (IJCV 2024 · similarity 0.90)
- Progressive Cognitive Human Parsing (AAAI 2018 · similarity 0.89)
- Instance-level Human Parsing via Part Grouping Network (ECCV 2018 · similarity 0.89)
- Devil in the Details: Towards Accurate Single and Multiple Human Parsing (AAAI 2019 · similarity 0.88)
- From Simple to Complex Scenes: Learning Robust Feature Representations for Accurate Human Parsing (TPAMI 2024 · 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.
