Unsupervised Person Re-identification by Combining Appearance Features with Spatialoral Features

Kunpeng Gao, Meibin Qi, Yonglai Wei, Zhiping Lai, Shaoqing Xu, Wei Wang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Most of the existing person re-identification methods usually follow a supervised learning framework and train models based on a large number of labeled pedestrian images. However, directly deploying these trained models in real scenes will lead to poor performances, because the target domain data may be completely different from the training data, thus the model parameters cannot be well fitted. Furthermore, it is very time-consuming and impractical to label a large number of data. In order to solve these problems, we propose a simple and effective strategy for segmentation based on key parts aiming to obtain the discriminative appearance features. Simultaneously, we constructs a hybrid Gaussian model by calculating the time difference of pedestrian groups to acquire spatialoral features. Finally, a measure fusion model is used to combine the appearance measure matrix and spatialoral distance matrix, which greatly improves the performance of the unsupervised person re-identification. We conduct extensive experiments on the large-scale image datasets, including Market-1501 and DukeMTMC-reID. The experimental results demonstrate that our algorithm is superior to state-of-the-art unsupervised re-identification approaches.

Original languageEnglish (US)
Title of host publicationProceedings of the 2020 5th International Conference on Multimedia Systems and Signal Processing, ICMSSP 2020
PublisherAssociation for Computing Machinery
Pages21-25
Number of pages5
ISBN (Electronic)9781450377485
DOIs
StatePublished - May 28 2020
Event5th International Conference on Multimedia Systems and Signal Processing, ICMSSP 2020 - Chengdu, China
Duration: May 28 2020May 30 2020

Publication series

NameACM International Conference Proceeding Series

Conference

Conference5th International Conference on Multimedia Systems and Signal Processing, ICMSSP 2020
Country/TerritoryChina
CityChengdu
Period05/28/2005/30/20

Keywords

  • Fusion model
  • Re-identification
  • Segmentation
  • Spatialoral features

ASJC Scopus subject areas

  • Software
  • Human-Computer Interaction
  • Computer Vision and Pattern Recognition
  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'Unsupervised Person Re-identification by Combining Appearance Features with Spatialoral Features'. Together they form a unique fingerprint.

Cite this