Research map / Label-efficient and robust learning / Recognition and long-tail
Camera-Aware Proxies for Unsupervised 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 |
|---|---|---|---|
| use camera-aware proxies and both intra- and inter-camera losses with proxy-balanced sampling instead of the baseline (ablation) | ResNet-50 Re-ID model · Market-1501 | Rank-1: 79.7 -> 91.4 | Helped |
| use camera-aware proxies and both intra- and inter-camera losses with proxy-balanced sampling instead of the baseline (ablation) | ResNet-50 Re-ID model · DukeMTMC-ReID | Rank-1: 74.3 -> 81.1 | Helped |
| use camera-aware proxies and both intra- and inter-camera losses with proxy-balanced sampling instead of the baseline (ablation) | ResNet-50 Re-ID model · MSMT17 | Rank-1: 34.0 -> 67.4 | Helped |
| add intra-camera loss to inter-camera loss (ablation) | CAP4 on MSMT17 · MSMT17 | Rank-1: 66.9 -> 67.4 | Helped |
| add intra-camera loss to inter-camera loss (ablation) | CAP4 on DukeMTMC-ReID · DukeMTMC-ReID | Rank-1: 78.0 -> 81.1 | Helped |
| add intra-camera loss to inter-camera loss (ablation) | CAP4 on Market-1501 · Market-1501 | Rank-1: 91.1 -> 91.4 | Helped |
| add inter-camera loss to intra-camera loss (ablation) | CAP2 · MSMT17 | Rank-1: 51.3 -> 67.4 | Helped |
| add proxy-balanced sampling instead of class-balanced sampling (ablation) | intra-camera-only CAP · MSMT17 | Rank-1: 48.6 -> 51.3 | Helped |
| add proxy-balanced sampling instead of class-balanced sampling (ablation) | inter-camera-only CAP · MSMT17 | Rank-1: 66.3 -> 66.9 | Helped |
| add proxy-balanced sampling instead of class-balanced sampling (ablation) | CAP with both losses · MSMT17 | Rank-1: 66.7 -> 67.4 | Helped |
| use intra-camera loss alone instead of the baseline objective (ablation) | CAP on Market-1501 · Market-1501 | mAP: 62.9 -> 58.9 | Hurt |
Closest papers by meaning
- Intra-Inter Camera Similarity for Unsupervised Person Re-Identification (CVPR 2021 · similarity 0.92)
- Joint Noise-Tolerant Learning and Meta Camera Shift Adaptation for Unsupervised Person Re-Identification (CVPR 2021 · similarity 0.91)
- Unsupervised Person Re-Identification via Softened Similarity Learning (CVPR 2020 · similarity 0.91)
- Exploring the Camera Bias of Person Re-identification (ICLR 2025 · similarity 0.88)
- Person De-reidentification: A Variation-guided Identity Shift Modeling (CVPR 2025 · similarity 0.88)
- Alice Benchmarks: Connecting Real World Re-Identification with the Synthetic (ICLR 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.