Research map / Label-efficient and robust learning / Recognition and long-tail
Dual Attention Matching Network for Context-Aware Feature Sequence Based 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 average pooling with dual attentive sequence matching (ablation) | DenseNet-121 person ReID · Market-1501 | R1: 74.20 -> 79.66 | Helped |
| replace average pooling with dual attentive sequence matching (ablation) | DenseNet-121 person ReID · DukeMTMC-reID | R1: 64.05 -> 68.40 | Helped |
| replace average pooling with dual attentive sequence matching (ablation) | DenseNet-121 person ReID · MARS | R1: 65.45 -> 66.36 | Helped |
| use intra-sequence attention only instead of dual attention (ablation) | DuATM · Market-1501 | R1: 79.66 -> 78.78 | Hurt |
| use inter-sequence attention only instead of dual attention (ablation) | DuATM · Market-1501 | R1: 79.66 -> 72.36 | Hurt |
| add decorrelation loss (ablation) | DuATM with triplet loss · Market-1501 | R1: 79.66 -> 81.83 | Helped |
| add decorrelation loss (ablation) | DuATM with triplet loss · DukeMTMC-reID | R1: 68.40 -> 69.17 | Helped |
| add decorrelation loss (ablation) | DuATM with triplet loss · MARS | R1: 66.36 -> 66.52 | Helped |
| add cross-entropy identity loss (ablation) | DuATM with triplet loss · Market-1501 | R1: 79.66 -> 87.50 | Helped |
| add cross-entropy identity loss (ablation) | DuATM with triplet loss · DukeMTMC-reID | R1: 68.40 -> 79.40 | Helped |
| add cross-entropy identity loss (ablation) | DuATM with triplet loss · MARS | R1: 66.36 -> 73.74 | Helped |
| increase decorrelation-loss weight beyond a moderate value (ablation) | DuATM · Market-1501, DukeMTMC-reID, and MARS | R1 | Hurt |
Closest papers by meaning
- Attention-Aware Compositional Network for Person Re-Identification (CVPR 2018 · similarity 0.89)
- Deep Neural Networks with Inexact Matching for Person Re-Identification (NeurIPS 2016 · similarity 0.88)
- Re-Identification With Consistent Attentive Siamese Networks (CVPR 2019 · similarity 0.88)
- Hierarchical Prompt Learning for Image- and Text-Based Person Re-Identification (AAAI 2026 · similarity 0.88)
- Single Camera Training for Person Re-Identification (AAAI 2020 · similarity 0.88)
- Cheb-GR: Rethinking K-nearest Neighbor Search in Re-ranking for Person Re-identification (CVPR 2025 · similarity 0.86)
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.