Research map / Image generation / Editing and style transfer
Feature Erasing and Diffusion Network for Occluded 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 NPO augmentation to the ViT baseline (ablation) | FED · Occluded-DukeMTMC | Rank-1: 59.1 -> 65.4 | Helped |
| Add random erasing to the ViT baseline (ablation) | FED · Occluded-DukeMTMC | Rank-1: 59.1 -> 60.3 | Helped |
| Add OEM to NPO augmentation (ablation) | FED · Occluded-DukeMTMC | mAP: 53.5 -> 55.4 | Helped |
| Add FDM to NPO augmentation (ablation) | FED · Occluded-DukeMTMC | Rank-1: 65.4 -> 67.1 | Helped |
| Add both OEM and FDM to NPO augmentation (ablation) | FED · Occluded-DukeMTMC | Rank-1: 65.4 -> 68.1 | Helped |
| Use a more diverse occlusion patch set, combining OS1 and OS2 (ablation) | FED · Occluded-REID | Rank-1: 86.3 -> 87.0 | Helped |
| Use a more diverse occlusion patch set, combining OS1 and OS2 (ablation) | FED · Partial-REID | mAP: 80.5 -> 82.3 | Helped |
| Increase memory-search K across the tested settings (ablation) | FED · Occluded-DukeMTMC | Rank-1: 67.4; 67.7; 67.3; 68.1 | Mixed |
| Use FED rather than TransReID | Person ReID · Occluded-DukeMTMC | Rank-1: 64.2 -> 68.1 | Helped |
| Use FED rather than TransReID | Person ReID · Market-1501 | mAP: 88.2 -> 86.3 | Hurt |
Closest papers by meaning
- Dynamic Feature Pruning and Consolidation for Occluded Person Re-identification (AAAI 2024 · similarity 0.90)
- Attention Disturbance and Dual-Path Constraint Network for Occluded Person Re-identification (AAAI 2024 · similarity 0.88)
- From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization (CVPR 2025 · similarity 0.88)
- Occluded Person Re-identification via Saliency-Guided Patch Transfer (AAAI 2024 · similarity 0.88)
- Text-Based Occluded Person Re-identification via Multi-Granularity Contrastive Consistency Learning (AAAI 2024 · similarity 0.88)
- Exploring the Camera Bias of Person Re-identification (ICLR 2025 · similarity 0.85)
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.