Research map / Detection and segmentation / CNN features
Sequential End-to-end Network for Efficient Person Search
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 |
|---|---|---|---|
| Replace parallel proposal-based processing with sequential detection followed by re-ID (ablation) | NAE person search network · CUHK-SYSU | mAP: 91.7 -> 92.5 | Helped |
| Replace parallel proposal-based processing with sequential detection followed by re-ID (ablation) | NAE person search network · PRW | mAP: 43.6 -> 44.7 | Helped |
| Add FCS using the first detection head's classification scores (ablation) | SeqNet · CUHK-SYSU | AP: 86.7->89.7 | Helped |
| Add FCS using the first detection head's classification scores (ablation) | SeqNet · CUHK-SYSU | mAP: 93.1→93.8 | Helped |
| Add NMS before the re-ID stage (ablation) | SeqNet · CUHK-SYSU | FPS: 7.4 to 11.5 | Helped |
| Add NMS before the re-ID stage (ablation) | SeqNet · CUHK-SYSU | mAP: 93.1->93.4 | No clear effect |
| Add the baseline-head regression branch alongside classification (ablation) | SeqNet baseline head · CUHK-SYSU | mAP: 93.4 to 93.8 | Helped |
| Add the baseline-head classification branch alongside regression (ablation) | SeqNet baseline head · CUHK-SYSU | mAP: 93.3 to 93.8 | Helped |
| Use detected BBoxes as input to the re-ID head instead of RPN proposals (ablation) | NAE re-identifier · CUHK-SYSU | mAP: 92.5 -> 93.3 | Helped |
| Use the SeqNet re-identifier instead of NAE re-identifier (ablation) | NAE detector · CUHK-SYSU | mAP: 92.5 -> 93.1 | Helped |
| Add CBGM context matching | NAE+SeqNet · CUHK-SYSU | mAP: 93.8 -> 94.8 | Helped |
| Add CBGM context matching | NAE+SeqNet · PRW | mAP: 46.7 -> 47.6 | Helped |
Closest papers by meaning
- Diverse Knowledge Distillation for End-to-End Person Search (AAAI 2021 · similarity 0.91)
- Query-Guided End-To-End Person Search (CVPR 2019 · similarity 0.90)
- Deep Neural Networks with Inexact Matching for Person Re-Identification (NeurIPS 2016 · similarity 0.89)
- Joint Detection and Identification Feature Learning for Person Search (CVPR 2017 · similarity 0.88)
- Cheb-GR: Rethinking K-nearest Neighbor Search in Re-ranking for Person Re-identification (CVPR 2025 · similarity 0.87)
- Divide and Conquer: Hybrid Pre-training for Person Search (AAAI 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.