Research map / Label-efficient and robust learning / Recognition and long-tail
Joint Learning of Single-Image and Cross-Image Representations for Person Re-Identification
Check 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 |
|---|---|---|---|
| Combine SIR and CIR matching scores in the pairwise model (ablation) | Pairwise comparison model · CUHK03 | rank-1 accuracy (%): 37.15 -> 43.36; 35.70 -> 43.36 | Helped |
| Combine SIR and CIR matching scores in the triplet model (ablation) | Triplet comparison model · CUHK03 | rank-1 accuracy (%): 43.23 -> 51.33; 43.46 -> 51.33 | Helped |
| Combine SIR and CIR matching scores in the pairwise-plus-triplet model (ablation) | Combined comparison model · CUHK03 | rank-1 accuracy (%): 44.35 -> 52.17; 45.40 -> 52.17 | Helped |
| Use triplet rather than pairwise comparison learning | Joint SIR-and-CIR model · CUHK03 | rank-1 accuracy (%): 43.36 -> 51.33 | Helped |
| Combine pairwise- and triplet-trained matching models | Joint SIR-and-CIR model · CUHK03 | rank-1 accuracy (%): 51.33 -> 52.17 | Helped |
| Pretrain on CUHK03 before fine-tuning | Pairwise comparison model · CUHK01 | rank-1 accuracy (%): 58.93% -> 61.20% | Helped |
| Pretrain on CUHK03 before fine-tuning | Triplet comparison model · CUHK01 | rank-1 accuracy (%): 65.21% -> 71.30% | Helped |
| Pretrain on CUHK03 before fine-tuning | Combined comparison model · CUHK01 | rank-1 accuracy (%): 65.71% -> 71.80% | Helped |
Closest papers by meaning
- Deep Neural Networks with Inexact Matching for Person Re-Identification (NeurIPS 2016 · similarity 0.92)
- Harmonious Attention Network for Person Re-Identification (CVPR 2018 · similarity 0.89)
- A Pose-Sensitive Embedding for Person Re-Identification With Expanded Cross Neighborhood Re-Ranking (CVPR 2018 · similarity 0.89)
- AXM-Net: Implicit Cross-Modal Feature Alignment for Person Re-identification (AAAI 2022 · similarity 0.89)
- Unsupervised Cross-Dataset Transfer Learning for Person Re-Identification (CVPR 2016 · similarity 0.89)
- Resource Aware Person Re-Identification Across Multiple Resolutions (CVPR 2018 · similarity 0.89)
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.