Research map / Label-efficient and robust learning / Recognition and long-tail
Identity-Guided Human Semantic Parsing for Person Re-Identification
Kuan Zhu, Haiyun Guo, Zhiwei Liu, Ming Tang, Jinqiao Wang
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 |
|---|---|---|---|
| use learned semantic parts (ISP) instead of extra semantics from SCHP (ablation) | ISP, HRNet-W32 · DukeMTMC-reID | mAP: 79.1 -> 80.0 | Improved |
| use learned semantic parts (ISP) instead of extra semantics from SCHP (ablation) | ISP, HRNet-W32 · CUHK03-NP Detected | mAP: 69.6 -> 71.4 | Improved |
| add extra SCHP semantics to the baseline (ablation) | HRNet-W32 baseline · DukeMTMC-reID | mAP: 77.2 -> 79.1 | Improved |
| add learned ISP parsing to the baseline (ablation) | HRNet-W32 baseline · CUHK03-NP Labeled | Rank-1: 71.9 -> 76.5 | Improved |
| add aligned representation matching with shared-visible parts (arm) (ablation) | ISP · Occluded-DukeMTMC | Rank-1: 59.5 -> 62.8 | Improved |
| add aligned representation matching with shared-visible parts (arm) (ablation) | ISP · Occluded-DukeMTMC | mAP: 51.4 -> 52.3 | Improved |
| use HRNet-W32 backbone instead of ResNet50 (ablation) | ISP · DukeMTMC-reID | mAP: 78.9 -> 80.0 | Improved |
| use HRNet-W32 backbone instead of SeResNet50 (ablation) | ISP · DukeMTMC-reID | Rank-1: 88.8 -> 89.6 | Improved |
| set number of parts K to 5 instead of 4 or 7 (ablation) | ISP · DukeMTMC-reID | Rank-1: 89.6 -> 88.6 | Hurt |
| vary number of parts K from 4 to 8 (ablation) | ISP · Market1501 | mAP: 88.6, 87.7, 88.4, 88.4, 88.6 for K=4..8 | No clear effect |
Closest papers by meaning
- Human Semantic Parsing for Person Re-Identification (CVPR 2018 · similarity 0.92)
- Beyond Human Parts: Dual Part-Aligned Representations for Person Re-Identification (ICCV 2019 · similarity 0.91)
- Semantics-Aligned Representation Learning for Person Re-Identification (AAAI 2020 · similarity 0.89)
- Densely Semantically Aligned Person Re-Identification (CVPR 2019 · similarity 0.89)
- Deeply-Learned Part-Aligned Representations for Person Re-Identification (ICCV 2017 · similarity 0.88)
- Part-Based Pseudo Label Refinement for Unsupervised Person Re-Identification (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.
