• Home
  • Search
  • Global Distance-Distributions Separation for Unsupervised Person Re-identification
  • Open Access IconOpen Access
  • Cite Icon74
  • https://doi.org/10.1007/978-3-030-58571-6_43Copy DOI Icon

Global Distance-Distributions Separation for Unsupervised Person Re-identification

  • Jan 1, 2020
  • Xin Jin +3 more
Show More
  • Abstract
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

Abstract Supervised person re-identification (ReID) often has poor scalability and usability in real-world deployments due to domain gaps and the lack of annotations for the target domain data. Unsupervised person ReID through domain adaptation is attractive yet challenging. Existing unsupervised ReID approaches often fail in correctly identifying the positive samples and negative samples through the distance-based matching/ranking. The two distributions of distances for positive sample pairs (Pos-distr) and negative sample pairs (Neg-distr) are often not well separated, having large overlap. To address this problem, we introduce a global distance-distributions separation (GDS) constraint over the two distributions to encourage the clear separation of positive and negative samples from a global view. We model the two global distance distributions as Gaussian distributions and push apart the two distributions while encouraging their sharpness in the unsupervised training process. Particularly, to model the distributions from a global view and facilitate the timely updating of the distributions and the GDS related losses, we leverage a momentum update mechanism for building and maintaining the distribution parameters (mean and variance) and calculate the loss on the fly during the training. Distribution-based hard mining is proposed to further promote the separation of the two distributions. We validate the effectiveness of the GDS constraint in unsupervised ReID networks. Extensive experiments on multiple ReID benchmark datasets show our method leads to significant improvement over the baselines and achieves the state-of-the-art performance.KeywordsUnsupervised learningPerson re-identificationGlobal distance-distributions separationMomentum updateHard mining

Similar Papers
  • Research Article

Camera Style Alignment for Unsupervised Person Re-Identification

  • Jan 01, 2025
  • IEICE Transactions on Information and Systems
  • Jifeng Guo +4
  • Conference Article
  • Citations40

Meta Pairwise Relationship Distillation for Unsupervised Person Re-identification

  • Oct 01, 2021
  • Haoxuanye Ji +5
  • Research Article
  • Citations39

Median Stable Clustering and Global Distance Classification for Cross-Domain Person Re-Identification

  • May 01, 2022
  • IEEE Transactions on Circuits and Systems for Video Technology
  • Zhiqi Pang +4
  • Research Article
  • Citations111

Hybrid Contrastive Learning for Unsupervised Person Re-Identification

  • Jan 01, 2023
  • IEEE Transactions on Multimedia
  • Tongzhen Si +3
  • Research Article
  • Citations2

Consistent positive correlation sample distribution: Alleviating the negative sample noise issue in contrastive adaptation

  • Nov 14, 2024
  • Expert Systems With Applications
  • Xing Wei +6
  • Book Chapter
  • Citations1

A Deep Clustering-Guide Learning for Unsupervised Person Re-identification

  • Jan 01, 2019
  • Guo Chen +2
  • Research Article
  • Citations9

Imitating targets from all sides: an unsupervised transfer learning method for person re-identification

  • Apr 02, 2021
  • International Journal of Machine Learning and Cybernetics
  • Jiajie Tian +4
  • Conference Article
  • Citations45

Hard-sample Guided Hybrid Contrast Learning for Unsupervised Person Re-Identification

  • Nov 17, 2021
  • Zheng Hu +2
  • Research Article
  • Citations3

Multiscale Subgraph Adversarial Contrastive Learning.

  • Aug 01, 2025
  • IEEE transactions on neural networks and learning systems
  • Yanbei Liu +6
  • Research Article
  • Citations133

Cluster-Guided Contrastive Graph Clustering Network

  • Jun 26, 2023
  • Proceedings of the AAAI Conference on Artificial Intelligence
  • Xihong Yang +8
  • Research Article
  • Citations4

Domain Adaptive Learning with Multi-Granularity Features for Unsupervised Person Re-identification

  • Jan 05, 2022
  • Chinese Journal of Electronics
  • Lihua Fu +5
  • Research Article

Silhouette-Driven Contrastive Learning with Hierarchical Neighborhood Structure for Unsupervised Clothes Change Person Re-Identification

  • Oct 01, 2025
  • CAAI Artificial Intelligence Research
  • Mingkun Li +1
  • Research Article
  • Citations192

Unsupervised Person Re-Identification by Deep Asymmetric Metric Embedding.

  • Dec 14, 2018
  • IEEE Transactions on Pattern Analysis and Machine Intelligence
  • Hong-Xing Yu +2
  • Research Article
  • Citations3

Frequency-Masked Embedding Inference: A Non-Contrastive Approach for Time Series Representation Learning

  • Apr 11, 2025
  • Proceedings of the AAAI Conference on Artificial Intelligence
  • En Fu +1
  • Research Article
  • Citations37

Viewpoint-Aware Progressive Clustering for Unsupervised Vehicle Re-Identification

  • Aug 01, 2022
  • IEEE Transactions on Intelligent Transportation Systems
  • Aihua Zheng +3
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.