• Home
  • Search
  • Improved Instance Discrimination and Feature Compactness for End-to-End Person Search
  • Cite Icon28
  • https://doi.org/10.1109/tcsvt.2021.3082775Copy DOI Icon

Improved Instance Discrimination and Feature Compactness for End-to-End Person Search

  • Abstract
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

Person search aims to locate and retrieve specific pedestrians in scene images, including two subtasks, pedestrian detection and person re-identification. Recently, triplet loss has been widely used in person re-identification, which effectively improves the pedestrian features embedding and achieves superior performance. However, forming triplet in the person search is not an easy task. Most of the existing end-to-end person search methods are based on Faster R-CNN. The training process of person re-identification part is affected by the detector. It is difficult to form pedestrian triplets within a limited batch size. Also, there are many pedestrian identities in the person search dataset, but each pedestrian identity only has a few samples. It is difficult to learn a robust pedestrian feature representation for person search. To resolve the problem discussed above, a novel Feature Compactness (FC) Loss for the person search is designed, which efficiently improves the inter-class discrimination and intra-class compactness of pedestrian features embedding without the need for positive or negative pairs. Besides, we propose a pedestrian attention module (PAM) to help the network focuses more on pedestrian information and suppresses irrelevant background information. Our method achieves comparable performance on two benchmarks, CUHK-SYSU and PRW, and achieves 91.96% of mAP and 93.34% of rank1 accuracy on CUHK-SYSU.

Similar Papers
  • Research Article
  • Citations15

Segmentation mask guided end-to-end person search

  • May 08, 2020
  • Signal Processing: Image Communication
  • Dingyuan Zheng +3
  • Research Article
  • Citations3

Perceiving informative key-points: A self-attention approach for person search

  • Nov 16, 2021
  • Signal Processing: Image Communication
  • Guangyu Gao +2
  • Research Article
  • Citations107

Sequential End-to-end Network for Efficient Person Search

  • May 18, 2021
  • Proceedings of the AAAI Conference on Artificial Intelligence
  • Zhengjia Li +1
  • Research Article
  • Citations5

End-to-end feature diversity person search with rank constraint of cross-class matrix

  • Nov 12, 2022
  • Neurocomputing
  • Yue Zhang +4
  • Research Article
  • Citations2

Person re-identification based on multi-appearance model

  • May 30, 2020
  • Multimedia Tools and Applications
  • Lei Huang +3
  • Book Chapter
  • Citations2

Person Search via Anchor-Free Detection and Part-Based Group Feature Similarity Estimation

  • Jan 01, 2020
  • Qing Liu +2
  • Conference Article
  • Citations10

Dhff: Robust Multi-Scale Person Search by Dynamic Hierarchical Feature Fusion

  • Sep 01, 2019
  • Yan Lu +4
  • Research Article
  • Citations312

HSME: Hypersphere Manifold Embedding for Visible Thermal Person Re-Identification

  • Jul 17, 2019
  • Proceedings of the AAAI Conference on Artificial Intelligence
  • Yi Hao +3
  • PDF
  • Research Article
  • Citations2

Enhancing Person Re-Identification through Attention-Driven Global Features and Angular Loss Optimization.

  • May 21, 2024
  • Entropy (Basel, Switzerland)
  • Yihan Bi +4
  • Conference Article
  • Citations5

Weakly Supervised Person Search

  • Oct 01, 2020
  • Lan Yan +3
  • Research Article
  • Citations20

Hybrid Modality Metric Learning for Visible-Infrared Person Re-Identification

  • Jan 25, 2022
  • ACM Transactions on Multimedia Computing, Communications, and Applications
  • La Zhang +8
  • Research Article

An Improved Split-Attention Architecture Based on Circle Loss for Person Re-Identification

  • Mar 31, 2021
  • International Journal of Intelligent Systems and Applications in Engineering
  • Zongjing Cao
  • Book Chapter

Cross Domain Person ReID Based on Symmetric Coding and Pedestrian Similarity

  • Jan 01, 2022
  • Xingyu Gao +3
  • Research Article
  • Citations25

PH-GCN: Person Retrieval With Part-Based Hierarchical Graph Convolutional Network

  • Jan 01, 2022
  • IEEE Transactions on Multimedia
  • Bo Jiang +4
  • Research Article
  • Citations6

An enhanced siamese angular softmax network with dual joint-attention for person re-identification

  • Feb 03, 2021
  • Applied Intelligence
  • Jie Su +4
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.