Fuzzy multilayer clustering and fuzzy label regularization for unsupervised person reidentification

Zhong Zhang*, Meiyan Huang, Shuang Liu, Baihua Xiao, Tariq S. Durrani

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

27 Citations (Scopus)
17 Downloads (Pure)

Abstract

Unsupervised person reidentification has received more attention due to its wide real-world applications. In this paper, we propose a novel method named fuzzy multilayer clustering (FMC) for unsupervised person reidentification. The proposed FMC learns a new feature space using a multilayer perceptron for clustering in order to overcome the influence of complex pedestrian images. Meanwhile, the proposed FMC generates fuzzy labels for unlabeled pedestrian images, which simultaneously considers the membership degree and the similarity between the sample and each cluster. We further propose the fuzzy label regularization (FLR) to train the convolutional neural network (CNN) using pedestrian images with fuzzy labels in a supervised manner. The proposed FLR could regularize the CNN training process and reduce the risk of overfitting. The effectiveness of our method is validated on three large-scale person reidentification databases, i.e., Market-1501, DukeMTMC-reID, and CUHK03.

Original languageEnglish
Pages (from-to)1356-1368
Number of pages13
JournalIEEE Transactions on Fuzzy Systems
Volume28
Issue number7
Early online date2 May 2019
DOIs
Publication statusPublished - 31 Jul 2020

Keywords

  • fuzzy label regularization (FLR)
  • fuzzy multilayer clustering (FMC)
  • unsupervised person reidentification

Fingerprint

Dive into the research topics of 'Fuzzy multilayer clustering and fuzzy label regularization for unsupervised person reidentification'. Together they form a unique fingerprint.

Cite this