COVAD: Content-Oriented Video Anomaly Detection using a Self-Attention based Deep Learning Model
Background: Video anomaly detection has always been a hot topic and attracting an increasing amount of attention. Much of the existing methods on video anomaly detection depend on processing the entire video rather than considering only the significant context. This paper proposes a novel video anom...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
KeAi Communications Co., Ltd.
2023-02-01
|
Series: | Virtual Reality & Intelligent Hardware |
Subjects: | |
Online Access: | http://www.sciencedirect.com/science/article/pii/S2096579622000481 |
_version_ | 1811161868551585792 |
---|---|
author | Wenhao Shao Praboda Rajapaksha Yanyan Wei Dun Li Noel Crespi Zhigang Luo |
author_facet | Wenhao Shao Praboda Rajapaksha Yanyan Wei Dun Li Noel Crespi Zhigang Luo |
author_sort | Wenhao Shao |
collection | DOAJ |
description | Background: Video anomaly detection has always been a hot topic and attracting an increasing amount of attention. Much of the existing methods on video anomaly detection depend on processing the entire video rather than considering only the significant context. This paper proposes a novel video anomaly detection method named COVAD, which mainly focuses on the region of interest in the video instead of the entire video. Our proposed COVAD method is based on an auto-encoded convolutional neural network and coordinated attention mechanism, which can effectively capture meaningful objects in the video and dependencies between different objects. Relying on the existing memory-guided video frame prediction network, our algorithm can more effectively predict the future motion and appearance of objects in the video. Our proposed algorithm obtained better experimental results on multiple data sets and outperformed the baseline models considered in our analysis. At the same time we improve a visual test that can provide pixel-level anomaly explanations. |
first_indexed | 2024-04-10T06:22:05Z |
format | Article |
id | doaj.art-d36d605bf12541c38703a865509159d0 |
institution | Directory Open Access Journal |
issn | 2096-5796 |
language | English |
last_indexed | 2024-04-10T06:22:05Z |
publishDate | 2023-02-01 |
publisher | KeAi Communications Co., Ltd. |
record_format | Article |
series | Virtual Reality & Intelligent Hardware |
spelling | doaj.art-d36d605bf12541c38703a865509159d02023-03-02T04:59:14ZengKeAi Communications Co., Ltd.Virtual Reality & Intelligent Hardware2096-57962023-02-01512441COVAD: Content-Oriented Video Anomaly Detection using a Self-Attention based Deep Learning ModelWenhao Shao0Praboda Rajapaksha1Yanyan Wei2Dun Li3Noel Crespi4Zhigang Luo5College of computer, National University of Defense Technology, Changsha 410073, CHIINA; Telecom SudParis, IMT, Institut Polytechnique de Paris, 91764 Palaiseau, FRANCE; Corresponding author.Telecom SudParis, IMT, Institut Polytechnique de Paris, 91764 Palaiseau, FRANCEZhengzhou College of Finance and Economics, 450000 Zhengzhou CHINATelecom SudParis, IMT, Institut Polytechnique de Paris, 91764 Palaiseau, FRANCETelecom SudParis, IMT, Institut Polytechnique de Paris, 91764 Palaiseau, FRANCECollege of computer, National University of Defense Technology, Changsha 410073, CHIINABackground: Video anomaly detection has always been a hot topic and attracting an increasing amount of attention. Much of the existing methods on video anomaly detection depend on processing the entire video rather than considering only the significant context. This paper proposes a novel video anomaly detection method named COVAD, which mainly focuses on the region of interest in the video instead of the entire video. Our proposed COVAD method is based on an auto-encoded convolutional neural network and coordinated attention mechanism, which can effectively capture meaningful objects in the video and dependencies between different objects. Relying on the existing memory-guided video frame prediction network, our algorithm can more effectively predict the future motion and appearance of objects in the video. Our proposed algorithm obtained better experimental results on multiple data sets and outperformed the baseline models considered in our analysis. At the same time we improve a visual test that can provide pixel-level anomaly explanations.http://www.sciencedirect.com/science/article/pii/S2096579622000481Video SurveillanceVideo Anomaly DetectionMachine LearningDeep LearningNeural NetworkCoordinate attention |
spellingShingle | Wenhao Shao Praboda Rajapaksha Yanyan Wei Dun Li Noel Crespi Zhigang Luo COVAD: Content-Oriented Video Anomaly Detection using a Self-Attention based Deep Learning Model Virtual Reality & Intelligent Hardware Video Surveillance Video Anomaly Detection Machine Learning Deep Learning Neural Network Coordinate attention |
title | COVAD: Content-Oriented Video Anomaly Detection using a Self-Attention based Deep Learning Model |
title_full | COVAD: Content-Oriented Video Anomaly Detection using a Self-Attention based Deep Learning Model |
title_fullStr | COVAD: Content-Oriented Video Anomaly Detection using a Self-Attention based Deep Learning Model |
title_full_unstemmed | COVAD: Content-Oriented Video Anomaly Detection using a Self-Attention based Deep Learning Model |
title_short | COVAD: Content-Oriented Video Anomaly Detection using a Self-Attention based Deep Learning Model |
title_sort | covad content oriented video anomaly detection using a self attention based deep learning model |
topic | Video Surveillance Video Anomaly Detection Machine Learning Deep Learning Neural Network Coordinate attention |
url | http://www.sciencedirect.com/science/article/pii/S2096579622000481 |
work_keys_str_mv | AT wenhaoshao covadcontentorientedvideoanomalydetectionusingaselfattentionbaseddeeplearningmodel AT prabodarajapaksha covadcontentorientedvideoanomalydetectionusingaselfattentionbaseddeeplearningmodel AT yanyanwei covadcontentorientedvideoanomalydetectionusingaselfattentionbaseddeeplearningmodel AT dunli covadcontentorientedvideoanomalydetectionusingaselfattentionbaseddeeplearningmodel AT noelcrespi covadcontentorientedvideoanomalydetectionusingaselfattentionbaseddeeplearningmodel AT zhigangluo covadcontentorientedvideoanomalydetectionusingaselfattentionbaseddeeplearningmodel |