Research map / Label-efficient and robust learning / Recognition and long-tail
Beyond Triplet Loss: A Deep Quadruplet Network 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 |
|---|---|---|---|
| add the different-probe constraint to the improved triplet loss (ablation) | Quadruplet network · CUHK03 | rank-1 accuracy: 72.78 -> 74.47 | Helped |
| add the different-probe constraint to the improved triplet loss (ablation) | Quadruplet network · CUHK01 (p=486) | rank-1 accuracy: 59.26 -> 62.55 | Helped |
| add the different-probe constraint to the improved triplet loss (ablation) | Quadruplet network · CUHK01 (p=100) | rank-1 accuracy: 78.00 -> 79.00 | Helped |
| add the different-probe constraint to the improved triplet loss (ablation) | Quadruplet network · VIPeR | rank-1 accuracy: 44.30 -> 48.42 | Helped |
| add softmax loss to the improved triplet network (ablation) | Triplet (Improved) · CUHK03 | rank-1 accuracy: 70.25 -> 72.78 | Helped |
| add softmax loss to the improved triplet network (ablation) | Triplet (Improved) · CUHK01 (p=486) | rank-1 accuracy: 58.85 -> 59.26 | Helped |
| add softmax loss to the improved triplet network (ablation) | Triplet (Improved) · CUHK01 (p=100) | rank-1 accuracy: 77.50 -> 78.00 | Helped |
| add softmax loss to the improved triplet network (ablation) | Triplet (Improved) · VIPeR | rank-1 accuracy: 44.30 -> 44.30 | No clear effect |
| add margin-based online hard negative mining (ablation) | Quadruplet · CUHK03 | rank-1 accuracy: 74.47 -> 75.53 | Helped |
| add margin-based online hard negative mining (ablation) | Quadruplet · CUHK01 (p=100) | rank-1 accuracy: 79.00 -> 81.00 | Helped |
| add margin-based online hard negative mining (ablation) | Quadruplet · VIPeR | rank-1 accuracy: 48.42 -> 49.05 | Helped |
| add margin-based online hard negative mining (ablation) | Quadruplet · VIPeR | rank-5 accuracy: 74.05 -> 73.10 | Hurt |
Closest papers by meaning
- Learning Incremental Triplet Margin for Person Re-Identification (AAAI 2019 · similarity 0.91)
- Person Re-Identification by Multi-Channel Parts-Based CNN With Improved Triplet Loss Function (CVPR 2016 · similarity 0.89)
- Single Camera Training for Person Re-Identification (AAAI 2020 · similarity 0.87)
- A Multi-Task Deep Network for Person Re-Identification (AAAI 2017 · similarity 0.87)
- Combined Depth Space Based Architecture Search for Person Re-Identification (CVPR 2021 · similarity 0.86)
- Instruct-ReID: A Multi-purpose Person Re-identification Task with Instructions (CVPR 2024 · similarity 0.84)
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.