Data-Augmentation for Reducing Dataset Bias in Person Re-identification

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Published

    View graph of relations

    In this paper we explore ways to address the issue of dataset bias in person re-identification by using data augmentation to increase the variability of the available datasets, and we introduce a novel data augmentation method for re-identification based on changing the image background. We show that use of data augmentation can improve the cross-dataset generalisation of convolutional network based re-identification systems, and that changing the image background yields further improvements.

    Documents

    • Data-Augmentation for Reducing Dataset Bias in Person Re-identification

      Rights statement: Copyright 2016 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other users, including reprinting/ republishing this material for advertising or promotional purposes, creating new collective works for resale or redistribution to servers or lists, or reuse of any copyrighted components of this work in other works.

      Accepted author manuscript, 3 MB, PDF-document

    DOI

    Original languageEnglish
    Title of host publicationProceedings of 2015 12th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS)
    PublisherInstitute of Electrical and Electronics Engineers (IEEE)
    Number of pages6
    ISBN (Print)978­1­4673­7632­7
    DOIs
    Publication statusPublished - Aug 2015
    Event 3rd AMMDS Workshop - AVSS 2015: Activity Monitoring by Multiple Distributed Sensing - Karlsruhe, Germany
    Duration: 25 Aug 2015 → …

    Workshop

    Workshop 3rd AMMDS Workshop - AVSS 2015: Activity Monitoring by Multiple Distributed Sensing
    CountryGermany
    CityKarlsruhe
    Period25/08/2015 → …

    ID: 16058766