A comprehensive study of sparse codes on abnormality detection

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

Research output: Contribution to conference without publisher/journalPaper without publisher/journalResearch

Abstract

Sparse representation has been applied successfully in abnor-mal event detection, in which the baseline is to learn a dic-tionary accompanied by sparse codes. While much empha-sis is put on discriminative dictionary construction, there areno comparative studies of sparse codes regarding abnormal-ity detection. We comprehensively study two types of sparsecodes solutions - greedy algorithms and convex L1-norm so-lutions - and their impact on abnormality detection perfor-mance. We also propose our framework of combining sparsecodes with different detection methods. Our comparative ex-periments are carried out from various angles to better un-derstand the applicability of sparse codes, including compu-tation time, reconstruction error, sparsity, detection accuracy,and their performance combining various detection methods.Experiments show that combining OMP codes with maxi-mum coordinate detection could achieve state-of-the-art per-formance on the UCSD dataset.
Original languageEnglish
Publication date13 Mar 2016
Publication statusPublished - 13 Mar 2016

Fingerprint

Error detection
Glossaries
Experiments

Cite this

@conference{1a52da694cff46f2bc1708d16335dea1,
title = "A comprehensive study of sparse codes on abnormality detection",
abstract = "Sparse representation has been applied successfully in abnor-mal event detection, in which the baseline is to learn a dic-tionary accompanied by sparse codes. While much empha-sis is put on discriminative dictionary construction, there areno comparative studies of sparse codes regarding abnormal-ity detection. We comprehensively study two types of sparsecodes solutions - greedy algorithms and convex L1-norm so-lutions - and their impact on abnormality detection perfor-mance. We also propose our framework of combining sparsecodes with different detection methods. Our comparative ex-periments are carried out from various angles to better un-derstand the applicability of sparse codes, including compu-tation time, reconstruction error, sparsity, detection accuracy,and their performance combining various detection methods.Experiments show that combining OMP codes with maxi-mum coordinate detection could achieve state-of-the-art per-formance on the UCSD dataset.",
author = "Huamin Ren and Hong Pan and Olsen, {S{\o}ren Ingvor} and Moeslund, {Thomas B.}",
year = "2016",
month = "3",
day = "13",
language = "English",
pages = "1--7",

}

A comprehensive study of sparse codes on abnormality detection. / Ren, Huamin; Pan, Hong; Olsen, Søren Ingvor; Moeslund, Thomas B.

2016. 1-7.

Research output: Contribution to conference without publisher/journalPaper without publisher/journalResearch

TY - CONF

T1 - A comprehensive study of sparse codes on abnormality detection

AU - Ren, Huamin

AU - Pan, Hong

AU - Olsen, Søren Ingvor

AU - Moeslund, Thomas B.

PY - 2016/3/13

Y1 - 2016/3/13

N2 - Sparse representation has been applied successfully in abnor-mal event detection, in which the baseline is to learn a dic-tionary accompanied by sparse codes. While much empha-sis is put on discriminative dictionary construction, there areno comparative studies of sparse codes regarding abnormal-ity detection. We comprehensively study two types of sparsecodes solutions - greedy algorithms and convex L1-norm so-lutions - and their impact on abnormality detection perfor-mance. We also propose our framework of combining sparsecodes with different detection methods. Our comparative ex-periments are carried out from various angles to better un-derstand the applicability of sparse codes, including compu-tation time, reconstruction error, sparsity, detection accuracy,and their performance combining various detection methods.Experiments show that combining OMP codes with maxi-mum coordinate detection could achieve state-of-the-art per-formance on the UCSD dataset.

AB - Sparse representation has been applied successfully in abnor-mal event detection, in which the baseline is to learn a dic-tionary accompanied by sparse codes. While much empha-sis is put on discriminative dictionary construction, there areno comparative studies of sparse codes regarding abnormal-ity detection. We comprehensively study two types of sparsecodes solutions - greedy algorithms and convex L1-norm so-lutions - and their impact on abnormality detection perfor-mance. We also propose our framework of combining sparsecodes with different detection methods. Our comparative ex-periments are carried out from various angles to better un-derstand the applicability of sparse codes, including compu-tation time, reconstruction error, sparsity, detection accuracy,and their performance combining various detection methods.Experiments show that combining OMP codes with maxi-mum coordinate detection could achieve state-of-the-art per-formance on the UCSD dataset.

M3 - Paper without publisher/journal

SP - 1

EP - 7

ER -