Research map / Label-efficient and robust learning / Recognition and long-tail
Cross-Modality Person Re-Identification via Modality Confusion and Center Aggregation
Xin Hao, Sanyuan Zhao, Mang Ye, Jianbing Shen
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 modality confusion module (ablation) | Base · SYSU-MM01 (all-search) | Rank-1 accuracy (%): 49.40 -> 51.46 | Improved |
| add modality confusion module (ablation) | AGW · SYSU-MM01 (all-search) | Rank-1 accuracy (%): 47.50 -> 49.29 | Improved |
| add modality confusion module (ablation) | AGW with random erasing · SYSU-MM01 (all-search) | Rank-1 accuracy (%): 59.82 -> 62.74 | Improved |
| add modality confusion module (ablation) | AGW with random erasing · SYSU-MM01 (all-search) | mAP (%): 56.07 -> 58.83 | Improved |
| add identity-aware marginal center aggregation (ICA) (ablation) | AGW with random erasing · SYSU-MM01 (all-search) | Rank-1 accuracy (%): 59.82 -> 63.56 | Improved |
| add camera-aware marginal center aggregation (CCA) (ablation) | AGW with random erasing · SYSU-MM01 (all-search) | Rank-1 accuracy (%): 59.82 -> 63.42 | Improved |
| add both ICA and CCA (ablation) | AGW with random erasing · SYSU-MM01 (all-search) | Rank-1 accuracy (%): 59.82 -> 64.13 | Improved |
| add both ICA and CCA (ablation) | Base · SYSU-MM01 (all-search) | Rank-1 accuracy (%): 49.40 -> 52.13 | Improved |
| add both ICA and CCA (ablation) | DDAG · SYSU-MM01 (all-search) | Rank-1 accuracy (%): 54.75 -> 57.27 | Improved |
| add modality confusion on top of ICA and CCA (ablation) | AGW with ICA and CCA · SYSU-MM01 (all-search) | Rank-1 accuracy (%): 64.13 -> 65.40 | Improved |
| set margin sigma to 8 rather than 0 (ablation) | MCLNet · SYSU-MM01 (all-search) | mINP (%): 45.68 -> 47.39 | Improved |
| use sigma 8 instead of larger margins such as 32 (ablation) | MCLNet · SYSU-MM01 (all-search) | Rank-1 accuracy (%): 64.29 -> 65.40 | Improved |
Closest papers by meaning
- Cross-Modality Person Re-Identification With Shared-Specific Feature Transfer (CVPR 2020 · similarity 0.92)
- Learning Memory-Augmented Unidirectional Metrics for Cross-Modality Person Re-Identification (CVPR 2022 · similarity 0.91)
- Interact, Embed, and EnlargE: Boosting Modality-Specific Representations for Multi-Modal Person Re-identification (AAAI 2022 · similarity 0.91)
- Modality Unifying Network for Visible-Infrared Person Re-Identification (ICCV 2023 · similarity 0.91)
- MDReID: Modality-Decoupled Learning for Any-to-Any Multi-Modal Object Re-Identification (NeurIPS 2025 · similarity 0.88)
- Cross-Modality Perturbation Synergy Attack for Person Re-identification (NeurIPS 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.
