Research map / Label-efficient and robust learning / Domain adaptation and generalization
Domain Adaptive Attention Learning for Unsupervised Person Re-Identification
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| replace hard pseudo-label loss with proposed soft-label loss (ablation) | pseudo-label Re-ID · Duke->Market | mAP: 61.1 -> 63.9 | Helped |
| replace LSR with proposed soft-label loss (ablation) | pseudo-label Re-ID · Duke->Market | mAP: 62.4 -> 63.9 | Helped |
| replace hard labels with soft-label loss (ablation) | pseudo-label Re-ID · Market->Duke | mAP: 56.2 -> 57.8 | Helped |
| add DAAM (residual attention) on top of DSH+DSP (ablation) | soft-label baseline · Duke->Market | mAP: 63.2 -> 67.8 | Helped |
| add DSP branch to DAAM + DSH (ablation) | soft-label baseline · Duke->Market | mAP: 64.7 -> 67.8 | Helped |
| use residual DAAM instead of two independent attention modules (ablation) | DSH+DSP model · Duke->Market | mAP: 66.6 -> 67.8 | Helped |
| add DSP branch without DAAM to soft-label baseline (ablation) | soft-label baseline · Duke->Market | mAP: 63.9 -> 63.2 | Hurt |
| add DSP branch without DAAM to soft-label baseline (ablation) | soft-label baseline · Market->Duke | mAP: 57.8 -> 59.7 | Helped |
| vary DBSCAN proportion p and min cluster size M (ablation) | full model · Duke->Market / Market->Duke | Rank-1 | No clear effect |
| more clustering iterations (1 to 5) (ablation) | full model · Duke->Market | mAP | Helped |
Closest papers by meaning
- Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identification (ICLR 2020 · similarity 0.93)
- Unsupervised Multi-Source Domain Adaptation for Person Re-Identification (CVPR 2021 · similarity 0.92)
- SECRET: Self-Consistent Pseudo Label Refinement for Unsupervised Domain Adaptive Person Re-identification (AAAI 2022 · similarity 0.90)
- Reliability Exploration with Self-Ensemble Learning for Domain Adaptive Person Re-identification (AAAI 2022 · similarity 0.90)
- Invariance Matters: Exemplar Memory for Domain Adaptive Person Re-Identification (CVPR 2019 · similarity 0.90)
- Group-aware Label Transfer for Domain Adaptive Person Re-identification (CVPR 2021 · similarity 0.90)
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.