Research map / Detection and segmentation / Weakly supervised detection and segmentation
Self-Correction for Human Parsing
Peike Li, Yunqiu Xu, Yunchao Wei, Yi Yang
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 model aggregation (ablation) | A-CE2P · LIP validation | mIoU: 56.88 -> 57.94 | Improved |
| add label refinement (ablation) | A-CE2P · LIP validation | mIoU: 56.88 -> 57.44 | Improved |
| add both model aggregation and label refinement (ablation) | A-CE2P · LIP validation | mIoU: 56.88 -> 58.62 | Improved |
| add edge loss (ablation) | CE2P-style parsing loss · LIP validation | mIoU: 53.75 -> 54.14 | Improved |
| add IoU loss (ablation) | A-CE2P · LIP validation | mIoU: 54.14 -> 56.33 | Improved |
| add consistency constraint (ablation) | A-CE2P · LIP validation | mIoU: 56.33 -> 56.88 | Improved |
| add SCHP to MobileNet-V2 backbone (ablation) | A-CE2P · LIP validation | mIoU: 52.06 -> 53.13 | Improved |
| add SCHP across backbones ResNet-18/50/101 (ablation) | A-CE2P · LIP validation | mIoU: +1.08, +1.40, +1.70 | Improved |
| add SCHP across context modules PSP/ASPP/OCNet (ablation) | A-CE2P · LIP validation | mIoU: +1.70, +1.30, +1.00 | Improved |
| add self-correction cycles (ablation) | SCHP · LIP validation | mIoU | Improved |
| replace CE2P with A-CE2P and add SCHP | CE2P · LIP validation | mIoU: 53.10 -> 58.62 | Improved |
| add test-time augmentation | A-CE2P w/ SCHP · LIP validation | mIoU: 58.62 -> 59.36 | Improved |
Closest papers by meaning
- Self-Learning With Rectification Strategy for Human Parsing (CVPR 2020 · similarity 0.89)
- Look Into Person: Self-Supervised Structure-Sensitive Learning and a New Benchmark for Human Parsing (CVPR 2017 · similarity 0.88)
- Semantic Human Parsing via Scalable Semantic Transfer Over Multiple Label Domains (CVPR 2023 · similarity 0.88)
- Deep Learning Technique for Human Parsing: A Survey and Outlook (IJCV 2024 · similarity 0.87)
- From Simple to Complex Scenes: Learning Robust Feature Representations for Accurate Human Parsing (TPAMI 2024 · 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.
