Multiple Granularities with Gradual Transition Network for Person Re-identification

Jialin Lu, Qingjie Zhao*, Lei Wang

*Corresponding author for this work

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

Abstract

Person re-identification (Re-ID) is a challenging task in computer vision, which aims at retrieving a target pedestrian from a gallery of person images captured from various cameras. Recent part-based methods, which employ horizontal splitting to integrate global and local information as final person representation, are not efficient enough in cases where the discriminative information near the splitting boundary is missing or incomplete due to partition. To address this issue, we proposed a novel method called Multiple Granularities with Gradual Transition Network (MGGTN) to fully mine fine-grained features at each part level and make the person representation more discriminative and robust. Our model introduces multi-branch network architecture to extract features with multiple granularities and uses a gradual transition strategy to obtain partial regions instead of easily partitioning the feature map into several stripes. Experimental results demonstrate the effectiveness of our method for Re-ID task. Especially, we achieve the new state-of-the-art results on both DukeMTMC-ReID and CUHK03 datasets and obtain the top rank1 result on Market1501 dataset.

Original languageEnglish
Title of host publicationCognitive Systems and Information Processing - 6th International Conference, ICCSIP 2021, Revised Selected Papers
EditorsFuchun Sun, Dewen Hu, Stefan Wermter, Lei Yang, Huaping Liu, Bin Fang
PublisherSpringer Science and Business Media Deutschland GmbH
Pages328-342
Number of pages15
ISBN (Print)9789811692468
DOIs
Publication statusPublished - 2022
Event6th International Conference on Cognitive Systems and Signal Processing, ICCSIP 2021 - Suzhou, China
Duration: 20 Nov 202121 Nov 2021

Publication series

NameCommunications in Computer and Information Science
Volume1515 CCIS
ISSN (Print)1865-0929
ISSN (Electronic)1865-0937

Conference

Conference6th International Conference on Cognitive Systems and Signal Processing, ICCSIP 2021
Country/TerritoryChina
CitySuzhou
Period20/11/2121/11/21

Keywords

  • Different granularities feature learning
  • Multi-branch network
  • Person re-identification

Fingerprint

Dive into the research topics of 'Multiple Granularities with Gradual Transition Network for Person Re-identification'. Together they form a unique fingerprint.

Cite this