Tri-modal Person Re-identification with RGB, Depth and Thermal Features

Andreas Møgelmose; Chris Bahnsen; Thomas B. Moeslund; Albert Clapés; Sergio Escalera

doi:10.1109/CVPRW.2013.52

Tri-modal Person Re-identification with RGB, Depth and Thermal Features

Andreas Møgelmose, Chris Bahnsen, Thomas B. Moeslund, Albert Clapés, Sergio Escalera

Research output: Contribution to book/anthology/report/conference proceeding › Article in proceeding › Research › peer-review

67 Citations (Scopus)

510 Downloads (Pure)

Abstract

Person re-identification is about recognizing people who have passed by a sensor earlier. Previous work is mainly based on RGB data, but in this work we for the first time present a system where we combine RGB, depth, and thermal data for re-identification purposes. First, from each of the three modalities, we obtain some particular features: from RGB data, we model color information from different regions of the body, from depth data, we compute different soft body biometrics, and from thermal data, we extract local structural information. Then, the three information types are combined in a joined classifier. The tri-modal system is evaluated on a new RGB-D-T dataset, showing successful results in re-identification scenarios.

Original language	English
Title of host publication	2013 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) : CVPRW 2013
Number of pages	7
Publisher	IEEE
Publication date	2013
Pages	301-307
ISBN (Print)	9781479909940
ISBN (Electronic)	978-0-7695-4990-3
DOIs	https://doi.org/10.1109/CVPRW.2013.52
Publication status	Published - 2013
Event	Perception Beyond the Visible Spectrum - Portland, United States Duration: 24 Jun 2013 → 24 Jun 2013 Conference number: 9

Workshop

Workshop	Perception Beyond the Visible Spectrum
Number	9
Country/Territory	United States
City	Portland
Period	24/06/2013 → 24/06/2013

Access to Document

10.1109/CVPRW.2013.52

triModalPersonReIdSubmitted manuscript, 2.19 MB

AUB Link

Search for the material in Aalborg University Library's search engine

AAU VAP Trimodal People Segmentation Dataset
Bahnsen, C. H. (Creator), Møgelmose, A. (Creator) & Moeslund, T. B. (Creator), Kaggle, 1 Jan 2017
DOI: 10.34740/kaggle/dsv/396257
Dataset: Supplementary material

Cite this

@inproceedings{73012690f17c4dc4b8805d398535c223,

title = "Tri-modal Person Re-identification with RGB, Depth and Thermal Features",

abstract = "Person re-identification is about recognizing people who have passed by a sensor earlier. Previous work is mainly based on RGB data, but in this work we for the first time present a system where we combine RGB, depth, and thermal data for re-identification purposes. First, from each of the three modalities, we obtain some particular features: from RGB data, we model color information from different regions of the body, from depth data, we compute different soft body biometrics, and from thermal data, we extract local structural information. Then, the three information types are combined in a joined classifier. The tri-modal system is evaluated on a new RGB-D-T dataset, showing successful results in re-identification scenarios.",

author = "Andreas M{\o}gelmose and Chris Bahnsen and Moeslund, {Thomas B.} and Albert Clap{\'e}s and Sergio Escalera",

year = "2013",

doi = "10.1109/CVPRW.2013.52",

language = "English",

isbn = "9781479909940",

pages = "301--307",

booktitle = "2013 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)",

publisher = "IEEE",

address = "United States",

note = "Perception Beyond the Visible Spectrum ; Conference date: 24-06-2013 Through 24-06-2013",

}

Møgelmose, A, Bahnsen, C, Moeslund, TB, Clapés, A & Escalera, S 2013, Tri-modal Person Re-identification with RGB, Depth and Thermal Features. in 2013 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW): CVPRW 2013. IEEE, pp. 301-307, Perception Beyond the Visible Spectrum, Portland, United States, 24/06/2013. https://doi.org/10.1109/CVPRW.2013.52

Tri-modal Person Re-identification with RGB, Depth and Thermal Features. / Møgelmose, Andreas; Bahnsen, Chris; Moeslund, Thomas B. et al.
2013 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW): CVPRW 2013. IEEE, 2013. p. 301-307.

Research output: Contribution to book/anthology/report/conference proceeding › Article in proceeding › Research › peer-review

TY - GEN

T1 - Tri-modal Person Re-identification with RGB, Depth and Thermal Features

AU - Møgelmose, Andreas

AU - Bahnsen, Chris

AU - Moeslund, Thomas B.

AU - Clapés, Albert

AU - Escalera, Sergio

N1 - Conference code: 9

PY - 2013

Y1 - 2013

N2 - Person re-identification is about recognizing people who have passed by a sensor earlier. Previous work is mainly based on RGB data, but in this work we for the first time present a system where we combine RGB, depth, and thermal data for re-identification purposes. First, from each of the three modalities, we obtain some particular features: from RGB data, we model color information from different regions of the body, from depth data, we compute different soft body biometrics, and from thermal data, we extract local structural information. Then, the three information types are combined in a joined classifier. The tri-modal system is evaluated on a new RGB-D-T dataset, showing successful results in re-identification scenarios.

AB - Person re-identification is about recognizing people who have passed by a sensor earlier. Previous work is mainly based on RGB data, but in this work we for the first time present a system where we combine RGB, depth, and thermal data for re-identification purposes. First, from each of the three modalities, we obtain some particular features: from RGB data, we model color information from different regions of the body, from depth data, we compute different soft body biometrics, and from thermal data, we extract local structural information. Then, the three information types are combined in a joined classifier. The tri-modal system is evaluated on a new RGB-D-T dataset, showing successful results in re-identification scenarios.

U2 - 10.1109/CVPRW.2013.52

DO - 10.1109/CVPRW.2013.52

M3 - Article in proceeding

SN - 9781479909940

SP - 301

EP - 307

BT - 2013 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

PB - IEEE

T2 - Perception Beyond the Visible Spectrum

Y2 - 24 June 2013 through 24 June 2013

ER -

Tri-modal Person Re-identification with RGB, Depth and Thermal Features

Abstract

Workshop

Access to Document

AUB Link

Fingerprint

Datasets

AAU VAP Trimodal People Segmentation Dataset

Cite this