Comparison of Multi-shot Models for Short-term Re-identification of People using RGB-D Sensors

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

1 Citation (Scopus)
360 Downloads (Pure)

Abstract

This work explores different types of multi-shot descriptors for re-identification in an on-the-fly enrolled environment using RGB-D sensors. We present a full re-identification pipeline complete with detection, segmentation, feature extraction, and re-identification, which expands on previous work by using multi-shot descriptors modeling people over a full camera pass instead of single frames with no temporal linking. We compare two different multi-shot models; mean histogram and histogram series, and test them each in 3 different color spaces. Both histogram descriptors are assisted by a depth-based pruning step where unlikely candidates are filtered away. Tests are run on 3 sequences captured in different circumstances and lighting situations to ensure proper generalization and lighting/environment invariance.
Original languageEnglish
Title of host publicationProceedings of the 10th International Conference on Computer Vision Theory and Applications (VISIGRAPP 2015)
EditorsJosé Braz, Sebastiano Battiato, Francisco Imai
Volume2
PublisherSCITEPRESS Digital Library
Publication date2015
Pages244-251
ISBN (Print)978-989-758-090-1
DOIs
Publication statusPublished - 2015
EventInternational Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications 2015 - Berlin, Germany
Duration: 11 Mar 201514 Mar 2015
Conference number: 11

Conference

ConferenceInternational Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications 2015
Number11
CountryGermany
CityBerlin
Period11/03/201514/03/2015

Fingerprint

Identification (control systems)
Lighting
Sensors
Invariance
Feature extraction
Pipelines
Cameras
Color

Cite this

Møgelmose, A., Bahnsen, C., & Moeslund, T. B. (2015). Comparison of Multi-shot Models for Short-term Re-identification of People using RGB-D Sensors. In J. Braz, S. Battiato, & F. Imai (Eds.), Proceedings of the 10th International Conference on Computer Vision Theory and Applications (VISIGRAPP 2015) (Vol. 2, pp. 244-251). SCITEPRESS Digital Library. https://doi.org/10.5220/0005266402440251
Møgelmose, Andreas ; Bahnsen, Chris ; Moeslund, Thomas B. / Comparison of Multi-shot Models for Short-term Re-identification of People using RGB-D Sensors. Proceedings of the 10th International Conference on Computer Vision Theory and Applications (VISIGRAPP 2015). editor / José Braz ; Sebastiano Battiato ; Francisco Imai. Vol. 2 SCITEPRESS Digital Library, 2015. pp. 244-251
@inproceedings{f1b0c6daa6a4406c96365cd36beeeb50,
title = "Comparison of Multi-shot Models for Short-term Re-identification of People using RGB-D Sensors",
abstract = "This work explores different types of multi-shot descriptors for re-identification in an on-the-fly enrolled environment using RGB-D sensors. We present a full re-identification pipeline complete with detection, segmentation, feature extraction, and re-identification, which expands on previous work by using multi-shot descriptors modeling people over a full camera pass instead of single frames with no temporal linking. We compare two different multi-shot models; mean histogram and histogram series, and test them each in 3 different color spaces. Both histogram descriptors are assisted by a depth-based pruning step where unlikely candidates are filtered away. Tests are run on 3 sequences captured in different circumstances and lighting situations to ensure proper generalization and lighting/environment invariance.",
author = "Andreas M{\o}gelmose and Chris Bahnsen and Moeslund, {Thomas B.}",
year = "2015",
doi = "10.5220/0005266402440251",
language = "English",
isbn = "978-989-758-090-1",
volume = "2",
pages = "244--251",
editor = "Jos{\'e} Braz and Sebastiano Battiato and Francisco Imai",
booktitle = "Proceedings of the 10th International Conference on Computer Vision Theory and Applications (VISIGRAPP 2015)",
publisher = "SCITEPRESS Digital Library",

}

Møgelmose, A, Bahnsen, C & Moeslund, TB 2015, Comparison of Multi-shot Models for Short-term Re-identification of People using RGB-D Sensors. in J Braz, S Battiato & F Imai (eds), Proceedings of the 10th International Conference on Computer Vision Theory and Applications (VISIGRAPP 2015). vol. 2, SCITEPRESS Digital Library, pp. 244-251, Berlin, Germany, 11/03/2015. https://doi.org/10.5220/0005266402440251

Comparison of Multi-shot Models for Short-term Re-identification of People using RGB-D Sensors. / Møgelmose, Andreas; Bahnsen, Chris; Moeslund, Thomas B.

Proceedings of the 10th International Conference on Computer Vision Theory and Applications (VISIGRAPP 2015). ed. / José Braz; Sebastiano Battiato; Francisco Imai. Vol. 2 SCITEPRESS Digital Library, 2015. p. 244-251.

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Comparison of Multi-shot Models for Short-term Re-identification of People using RGB-D Sensors

AU - Møgelmose, Andreas

AU - Bahnsen, Chris

AU - Moeslund, Thomas B.

PY - 2015

Y1 - 2015

N2 - This work explores different types of multi-shot descriptors for re-identification in an on-the-fly enrolled environment using RGB-D sensors. We present a full re-identification pipeline complete with detection, segmentation, feature extraction, and re-identification, which expands on previous work by using multi-shot descriptors modeling people over a full camera pass instead of single frames with no temporal linking. We compare two different multi-shot models; mean histogram and histogram series, and test them each in 3 different color spaces. Both histogram descriptors are assisted by a depth-based pruning step where unlikely candidates are filtered away. Tests are run on 3 sequences captured in different circumstances and lighting situations to ensure proper generalization and lighting/environment invariance.

AB - This work explores different types of multi-shot descriptors for re-identification in an on-the-fly enrolled environment using RGB-D sensors. We present a full re-identification pipeline complete with detection, segmentation, feature extraction, and re-identification, which expands on previous work by using multi-shot descriptors modeling people over a full camera pass instead of single frames with no temporal linking. We compare two different multi-shot models; mean histogram and histogram series, and test them each in 3 different color spaces. Both histogram descriptors are assisted by a depth-based pruning step where unlikely candidates are filtered away. Tests are run on 3 sequences captured in different circumstances and lighting situations to ensure proper generalization and lighting/environment invariance.

U2 - 10.5220/0005266402440251

DO - 10.5220/0005266402440251

M3 - Article in proceeding

SN - 978-989-758-090-1

VL - 2

SP - 244

EP - 251

BT - Proceedings of the 10th International Conference on Computer Vision Theory and Applications (VISIGRAPP 2015)

A2 - Braz, José

A2 - Battiato, Sebastiano

A2 - Imai, Francisco

PB - SCITEPRESS Digital Library

ER -

Møgelmose A, Bahnsen C, Moeslund TB. Comparison of Multi-shot Models for Short-term Re-identification of People using RGB-D Sensors. In Braz J, Battiato S, Imai F, editors, Proceedings of the 10th International Conference on Computer Vision Theory and Applications (VISIGRAPP 2015). Vol. 2. SCITEPRESS Digital Library. 2015. p. 244-251 https://doi.org/10.5220/0005266402440251