The University of Southampton
University of Southampton Institutional Repository

On Acquisition and Analysis of a Dataset Comprising of Gait, Ear and Semantic data

On Acquisition and Analysis of a Dataset Comprising of Gait, Ear and Semantic data
On Acquisition and Analysis of a Dataset Comprising of Gait, Ear and Semantic data
In outdoor scenarios such as surveillance where there is very little control over the environments, complex computer vision algorithms are often required for analysis. However constrained environments, such as walkways in airports where the surroundings and the path taken by individuals can be controlled, provide an ideal application for such systems. Figure 1.1 depicts an idealised constrained environment. The path taken by the subject is restricted to a narrow path and once inside is in a volume where lighting and other conditions are controlled to facilitate biometric analysis. The ability to control the surroundings and the flow of people greatly simplifes the computer vision task, compared to typical unconstrained environments. Even though biometric datasets with greater than one hundred people are increasingly common, there is still very little known about the inter and intra-subject variation in many biometrics. This information is essential to estimate the recognition capability and limits of automatic recognition systems. In order to accurately estimate the inter- and the intra- class variance, substantially larger datasets are required [40]. Covariates such as facial expression, headwear, footwear type, surface type and carried items are attracting increasing attention; although considering the potentially large impact on an individuals biometrics, large trials need to be conducted to establish how much variance results. This chapter is the first description of the multibiometric data acquired using the University of Southampton's Multi-Biometric Tunnel [26, 37]; a biometric portal using automatic gait, face and ear recognition for identification purposes. The tunnel provides a constrained environment and is ideal for use in high throughput security scenarios and for the collection of large datasets. We describe the current state of data acquisition of face, gait, ear, and semantic data and present early results showing the quality and range of data that has been collected. The main novelties of this dataset in comparison with other multi-biometric datasets are: 1. gait data exists for multiple views and is synchronised, allowing 3D reconstruction and analysis; 2. the face data is a sequence of images allowing for face recognition in video; 3. the ear data is acquired in a relatively unconstrained environment, as a subject walks past; and 4. the semantic data is considerably more extensive than has been available previously. We shall aim to show the advantages of this new data in biometric analysis, though the scope for such analysis is considerably greater than time and space allows for here.
9780521115964
277-301
CUP
Samangooei, Sina
c380fb26-55d4-4b34-94e7-c92bbb26a40d
Bustard, John
8fa23e3b-8594-4c87-81c4-a77e774c8c0b
Seely, Richard
0b26e904-a950-4041-8234-3a6e2bea4a91
Nixon, Mark
2b5b9804-5a81-462a-82e6-92ee5fa74e12
Carter, John
e05be2f9-991d-4476-bb50-ae91606389da
Samangooei, Sina
c380fb26-55d4-4b34-94e7-c92bbb26a40d
Bustard, John
8fa23e3b-8594-4c87-81c4-a77e774c8c0b
Seely, Richard
0b26e904-a950-4041-8234-3a6e2bea4a91
Nixon, Mark
2b5b9804-5a81-462a-82e6-92ee5fa74e12
Carter, John
e05be2f9-991d-4476-bb50-ae91606389da

Samangooei, Sina, Bustard, John, Seely, Richard, Nixon, Mark and Carter, John (2011) On Acquisition and Analysis of a Dataset Comprising of Gait, Ear and Semantic data. In, Multibiometrics for Human Identification. CUP, pp. 277-301.

Record type: Book Section

Abstract

In outdoor scenarios such as surveillance where there is very little control over the environments, complex computer vision algorithms are often required for analysis. However constrained environments, such as walkways in airports where the surroundings and the path taken by individuals can be controlled, provide an ideal application for such systems. Figure 1.1 depicts an idealised constrained environment. The path taken by the subject is restricted to a narrow path and once inside is in a volume where lighting and other conditions are controlled to facilitate biometric analysis. The ability to control the surroundings and the flow of people greatly simplifes the computer vision task, compared to typical unconstrained environments. Even though biometric datasets with greater than one hundred people are increasingly common, there is still very little known about the inter and intra-subject variation in many biometrics. This information is essential to estimate the recognition capability and limits of automatic recognition systems. In order to accurately estimate the inter- and the intra- class variance, substantially larger datasets are required [40]. Covariates such as facial expression, headwear, footwear type, surface type and carried items are attracting increasing attention; although considering the potentially large impact on an individuals biometrics, large trials need to be conducted to establish how much variance results. This chapter is the first description of the multibiometric data acquired using the University of Southampton's Multi-Biometric Tunnel [26, 37]; a biometric portal using automatic gait, face and ear recognition for identification purposes. The tunnel provides a constrained environment and is ideal for use in high throughput security scenarios and for the collection of large datasets. We describe the current state of data acquisition of face, gait, ear, and semantic data and present early results showing the quality and range of data that has been collected. The main novelties of this dataset in comparison with other multi-biometric datasets are: 1. gait data exists for multiple views and is synchronised, allowing 3D reconstruction and analysis; 2. the face data is a sequence of images allowing for face recognition in video; 3. the ear data is acquired in a relatively unconstrained environment, as a subject walks past; and 4. the semantic data is considerably more extensive than has been available previously. We shall aim to show the advantages of this new data in biometric analysis, though the scope for such analysis is considerably greater than time and space allows for here.

Text
multibioHID.pdf - Author's Original
Download (4MB)

More information

Published date: June 2011
Additional Information: Chapter: 12
Organisations: Vision, Learning and Control

Identifiers

Local EPrints ID: 272952
URI: http://eprints.soton.ac.uk/id/eprint/272952
ISBN: 9780521115964
PURE UUID: 15c6d1a2-a85e-469b-a05b-104b6d84e533
ORCID for Mark Nixon: ORCID iD orcid.org/0000-0002-9174-5934

Catalogue record

Date deposited: 20 Oct 2011 16:50
Last modified: 30 Sep 2020 16:32

Export record

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of http://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×