An In-depth Study of Sparse Codes on Abnormality Detection

Huamin Ren, Hong Pan, Søren Ingvor Olsen, Morten Bornø Jensen, Thomas B. Moeslund

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

1 Citation (Scopus)
274 Downloads (Pure)

Abstract

Sparse representation has been applied successfully in
abnormal event detection, in which the baseline is to learn
a dictionary accompanied by sparse codes. While much
emphasis is put on discriminative dictionary construction,
there are no comparative studies of sparse codes regarding abnormality detection. We present an in-depth study of
two types of sparse codes solutions - greedy algorithms and
convex L1-norm solutions - and their impact on abnormal-
ity detection performance. We also propose our framework
of combining sparse codes with different detection methods.
Our comparative experiments are carried out from various angles to better understand the applicability of sparse
codes, including computation time, reconstruction error,
sparsity, detection accuracy, and their performance combining various detection methods. The experiment results
show that combining OMP codes with maximum coordinate
detection could achieve state-of-the-art performance on the
UCSD dataset [14].
Original languageEnglish
Title of host publicationIEEE Advanced Video and Signal-based Surveillance (AVSS) 2016
Number of pages7
PublisherIEEE
Publication date2016
Pages1-7
ISBN (Print)978-1-5090-3812-1
ISBN (Electronic)978-1-5090-3811-4
DOIs
Publication statusPublished - 2016
EventThe 13th International Conference on Advanced Video and Signal-Based Surveillance - Colorado Springs, United States
Duration: 23 Aug 201626 Aug 2016
Conference number: 13
http://avss2016.org/

Conference

ConferenceThe 13th International Conference on Advanced Video and Signal-Based Surveillance
Number13
CountryUnited States
City Colorado Springs
Period23/08/201626/08/2016
Internet address

Fingerprint

Glossaries
Error detection
Experiments

Bibliographical note

Sparse Codes, Abnormality Detection

Cite this

Ren, H., Pan, H., Olsen, S. I., Jensen, M. B., & Moeslund, T. B. (2016). An In-depth Study of Sparse Codes on Abnormality Detection. In IEEE Advanced Video and Signal-based Surveillance (AVSS) 2016 (pp. 1-7). IEEE. https://doi.org/10.1109/AVSS.2016.7738016
Ren, Huamin ; Pan, Hong ; Olsen, Søren Ingvor ; Jensen, Morten Bornø ; Moeslund, Thomas B. / An In-depth Study of Sparse Codes on Abnormality Detection. IEEE Advanced Video and Signal-based Surveillance (AVSS) 2016. IEEE, 2016. pp. 1-7
@inproceedings{adf61ca212ab4ae58570ded8dbad8137,
title = "An In-depth Study of Sparse Codes on Abnormality Detection",
abstract = "Sparse representation has been applied successfully inabnormal event detection, in which the baseline is to learna dictionary accompanied by sparse codes. While muchemphasis is put on discriminative dictionary construction,there are no comparative studies of sparse codes regarding abnormality detection. We present an in-depth study oftwo types of sparse codes solutions - greedy algorithms andconvex L1-norm solutions - and their impact on abnormal-ity detection performance. We also propose our frameworkof combining sparse codes with different detection methods.Our comparative experiments are carried out from various angles to better understand the applicability of sparsecodes, including computation time, reconstruction error,sparsity, detection accuracy, and their performance combining various detection methods. The experiment resultsshow that combining OMP codes with maximum coordinatedetection could achieve state-of-the-art performance on theUCSD dataset [14].",
author = "Huamin Ren and Hong Pan and Olsen, {S{\o}ren Ingvor} and Jensen, {Morten Born{\o}} and Moeslund, {Thomas B.}",
note = "Sparse Codes, Abnormality Detection",
year = "2016",
doi = "10.1109/AVSS.2016.7738016",
language = "English",
isbn = "978-1-5090-3812-1",
pages = "1--7",
booktitle = "IEEE Advanced Video and Signal-based Surveillance (AVSS) 2016",
publisher = "IEEE",
address = "United States",

}

Ren, H, Pan, H, Olsen, SI, Jensen, MB & Moeslund, TB 2016, An In-depth Study of Sparse Codes on Abnormality Detection. in IEEE Advanced Video and Signal-based Surveillance (AVSS) 2016. IEEE, pp. 1-7, The 13th International Conference on Advanced Video and Signal-Based Surveillance, Colorado Springs, United States, 23/08/2016. https://doi.org/10.1109/AVSS.2016.7738016

An In-depth Study of Sparse Codes on Abnormality Detection. / Ren, Huamin; Pan, Hong; Olsen, Søren Ingvor; Jensen, Morten Bornø; Moeslund, Thomas B.

IEEE Advanced Video and Signal-based Surveillance (AVSS) 2016. IEEE, 2016. p. 1-7.

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

TY - GEN

T1 - An In-depth Study of Sparse Codes on Abnormality Detection

AU - Ren, Huamin

AU - Pan, Hong

AU - Olsen, Søren Ingvor

AU - Jensen, Morten Bornø

AU - Moeslund, Thomas B.

N1 - Sparse Codes, Abnormality Detection

PY - 2016

Y1 - 2016

N2 - Sparse representation has been applied successfully inabnormal event detection, in which the baseline is to learna dictionary accompanied by sparse codes. While muchemphasis is put on discriminative dictionary construction,there are no comparative studies of sparse codes regarding abnormality detection. We present an in-depth study oftwo types of sparse codes solutions - greedy algorithms andconvex L1-norm solutions - and their impact on abnormal-ity detection performance. We also propose our frameworkof combining sparse codes with different detection methods.Our comparative experiments are carried out from various angles to better understand the applicability of sparsecodes, including computation time, reconstruction error,sparsity, detection accuracy, and their performance combining various detection methods. The experiment resultsshow that combining OMP codes with maximum coordinatedetection could achieve state-of-the-art performance on theUCSD dataset [14].

AB - Sparse representation has been applied successfully inabnormal event detection, in which the baseline is to learna dictionary accompanied by sparse codes. While muchemphasis is put on discriminative dictionary construction,there are no comparative studies of sparse codes regarding abnormality detection. We present an in-depth study oftwo types of sparse codes solutions - greedy algorithms andconvex L1-norm solutions - and their impact on abnormal-ity detection performance. We also propose our frameworkof combining sparse codes with different detection methods.Our comparative experiments are carried out from various angles to better understand the applicability of sparsecodes, including computation time, reconstruction error,sparsity, detection accuracy, and their performance combining various detection methods. The experiment resultsshow that combining OMP codes with maximum coordinatedetection could achieve state-of-the-art performance on theUCSD dataset [14].

U2 - 10.1109/AVSS.2016.7738016

DO - 10.1109/AVSS.2016.7738016

M3 - Article in proceeding

SN - 978-1-5090-3812-1

SP - 1

EP - 7

BT - IEEE Advanced Video and Signal-based Surveillance (AVSS) 2016

PB - IEEE

ER -

Ren H, Pan H, Olsen SI, Jensen MB, Moeslund TB. An In-depth Study of Sparse Codes on Abnormality Detection. In IEEE Advanced Video and Signal-based Surveillance (AVSS) 2016. IEEE. 2016. p. 1-7 https://doi.org/10.1109/AVSS.2016.7738016