Research map / Label-efficient and robust learning / Recognition and long-tail
Person Re-Identification by Multi-Channel Parts-Based CNN With Improved Triplet Loss Function
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 multi-channel full-body and body-part features (ablation) | person re-identification CNN with original triplet loss · i-LIDS | Top1: 43.2 -> 57.2 | Helped |
| use the improved triplet loss (ablation) | multi-channel person re-identification CNN · i-LIDS | Top1: 57.2 -> 60.4 | Helped |
| add the multi-channel full-body and body-part features (ablation) | person re-identification CNN with original triplet loss · PRID2011 | Top1: 17.0 -> 22.0 | Helped |
| use the improved triplet loss (ablation) | multi-channel person re-identification CNN · PRID2011 | Top1: 22.0 -> 22.0 | No clear effect |
| add the multi-channel full-body and body-part features (ablation) | person re-identification CNN with original triplet loss · VIPeR | Top1: 34.3 -> 43.8 | Helped |
| use the improved triplet loss (ablation) | multi-channel person re-identification CNN · VIPeR | Top1: 43.8 -> 47.8 | Helped |
| add the multi-channel full-body and body-part features (ablation) | person re-identification CNN with original triplet loss · CUHK01 | Top1: 46.0 -> 52.3 | Helped |
| use the improved triplet loss (ablation) | multi-channel person re-identification CNN · CUHK01 | Top1: 52.3 -> 53.7 | Helped |
| add the intra-class distance constraint through improved triplet loss (ablation) | single-channel CNN · i-LIDS | Top1: 43.2 -> 47.3 | Helped |
| add the intra-class distance constraint through improved triplet loss (ablation) | single-channel CNN · PRID2011 | Top1: 17.0 -> 15.0 | Hurt |
| add the intra-class distance constraint through improved triplet loss (ablation) | single-channel CNN · VIPeR | Top1: 34.3 -> 37.2 | Helped |
| add the intra-class distance constraint through improved triplet loss (ablation) | single-channel CNN · CUHK01 | Top1: 46.0 -> 49.3 | Helped |
Closest papers by meaning
- Joint Learning of Single-Image and Cross-Image Representations for Person Re-Identification (CVPR 2016 · similarity 0.89)
- Beyond Triplet Loss: A Deep Quadruplet Network for Person Re-Identification (CVPR 2017 · similarity 0.89)
- Person Re-Identification With Cascaded Pairwise Convolutions (CVPR 2018 · similarity 0.88)
- Harmonious Attention Network for Person Re-Identification (CVPR 2018 · similarity 0.88)
- Interaction-And-Aggregation Network for Person Re-Identification (CVPR 2019 · similarity 0.88)
- Deep Neural Networks with Inexact Matching for Person Re-Identification (NeurIPS 2016 · 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.