Multi-view gait fusion for large scale human identification in surveillance videos

Emdad Hossain, Girija Chetty

Research output: A Conference proceeding or a Chapter in BookConference contribution

Abstract

In this paper we propose a novel multi-view feature fusion of gait biometric information in surveillance videos for large scale human identification. The experimental evaluation on low resolution surveillance video images from a publicly available database [1] showed that the combined LDA-MLP technique turns out to be a powerful method for capturing identity specific information from walking gait patterns. The multi-view fusion at feature level allows complementarity of multiple camera views in surveillance scenarios to be exploited for improvement of identity recognition performance.
Original languageEnglish
Title of host publicationInternational Conference on Advanced Concepts in intelligent Vision Systems (ACIVS 2012)
Subtitle of host publicationLecture Notes in Computer Science
EditorsJacques Blanc-Talon, Wilfried Philips, Dan Popescu, Paul Scheunders, Pavel Zemcik
Place of PublicationCzech Republic
PublisherSpringer
Pages527-538
Number of pages12
Volume7517
ISBN (Electronic)9783642331404
ISBN (Print)9783642331398
DOIs
Publication statusPublished - 2012
EventAdvances Concepts for Intelligent Vision System - Brno, Brno, Czech Republic
Duration: 4 Sep 2012 → …

Conference

ConferenceAdvances Concepts for Intelligent Vision System
CountryCzech Republic
CityBrno
Period4/09/12 → …

    Fingerprint

Cite this

Hossain, E., & Chetty, G. (2012). Multi-view gait fusion for large scale human identification in surveillance videos. In J. Blanc-Talon, W. Philips, D. Popescu, P. Scheunders, & P. Zemcik (Eds.), International Conference on Advanced Concepts in intelligent Vision Systems (ACIVS 2012): Lecture Notes in Computer Science (Vol. 7517, pp. 527-538). Czech Republic: Springer. https://doi.org/10.1007/978-3-642-33140-4_46