Research map / Label-efficient and robust learning / Recognition and long-tail
Matching on Sets: Conquer Occluded Person Re-identification Without Alignment
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 |
|---|---|---|---|
| match pattern sets with Jaccard similarity and add Jaccard triplet loss plus conflict penalty | MoS · Occluded-DukeMTMC | Rank-1 | Helped |
| match pattern sets with Jaccard similarity and add Jaccard triplet loss plus conflict penalty | MoS · P-ETHZ | Rank-1 | Helped |
| match pattern sets with Jaccard similarity and add Jaccard triplet loss plus conflict penalty | MoS · Market-1501 | mAP | Helped |
| match pattern sets with Jaccard similarity and add Jaccard triplet loss plus conflict penalty | MoS · DukeMTMC-reID | mAP | Helped |
| replace Euclidean similarity with Jaccard similarity (ablation) | Euclidean-triplet re-ID model · Occluded-DukeMTMC | mAP: 49.5 -> 50.8 | Helped |
| replace Euclidean triplet loss with Jaccard triplet loss (ablation) | set-matching model using Jaccard similarity · Occluded-DukeMTMC | Rank-1: 60.3 -> 64.4 | Helped |
| replace Euclidean triplet loss with Jaccard triplet loss (ablation) | set-matching model using Jaccard similarity · Occluded-DukeMTMC | mAP: 50.8 -> 54.8 | Helped |
| add pattern conflict penalty at inference (ablation) | Jaccard-triplet set-matching model · Occluded-DukeMTMC | Rank-1: 64.4 -> 66.6 | Helped |
| add pattern conflict penalty at inference (ablation) | Jaccard-triplet set-matching model · Occluded-DukeMTMC | mAP: 54.8 -> 55.1 | Helped |
| use cosine instead of Euclidean similarity (ablation) | Euclidean-triplet re-ID model · Occluded-DukeMTMC | Rank-1: 60.2 -> 58.6 | Hurt |
| vary the smoothing factor τ (ablation) | MoS · Occluded-DukeMTMC | Rank-1 | Mixed |
Closest papers by meaning
- Dynamic Feature Pruning and Consolidation for Occluded Person Re-identification (AAAI 2024 · similarity 0.89)
- Pose-Guided Feature Disentangling for Occluded Person Re-identification Based on Transformer (AAAI 2022 · similarity 0.89)
- High-Order Information Matters: Learning Relation and Topology for Occluded Person Re-Identification (CVPR 2020 · similarity 0.89)
- Deep Neural Networks with Inexact Matching for Person Re-Identification (NeurIPS 2016 · similarity 0.89)
- Deep Spatial Feature Reconstruction for Partial Person Re-Identification: Alignment-Free Approach (CVPR 2018 · similarity 0.88)
- CA-Jaccard: Camera-aware Jaccard Distance for Person Re-identification (CVPR 2024 · similarity 0.87)
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.