File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1117/12.2543564
- Scopus: eid_2-s2.0-85077816141
- WOS: WOS:000511402300093
- Find via
Supplementary
- Citations:
- Appears in Collections:
Conference Paper: Video-based Violence Detection by Human Action Analysis with Neural Network
Title | Video-based Violence Detection by Human Action Analysis with Neural Network |
---|---|
Authors | |
Keywords | Human action analysis Long-Short-Term-Memory Neural network Pose estimation Residual learning Video processing Violence detection |
Issue Date | 2019 |
Publisher | SPIE - International Society for Optical Engineering. The Journal's web site is located at http://spie.org/x1848.xml?WT.svl=mddp2 |
Citation | 2nd International conference on Image, Video Processing and Artificial Intelligence (IVPAI2019), Shanghai, China, 23-25 August 2019. In Proceedings of SPIE, 2019, v. 11321, p. 113212N:1-8 How to Cite? |
Abstract | In recent years, human action analysis is a focal point in video processing, especially on action recognition and safety surveillance. It always performs as an auxiliary tool to minimize the manpower-resource on special tasks. This paper explores the human action analysis in a specified situation, based on the human posture extraction by pose-estimation algorithm. Deep neural network (DNN) methods was used, composed of residual learning blocks for feature extraction and recurrent neural network for time-series data learning. All these modules can be applied on real-time videos, classifying different security levels of actions between two people, with 91.8% accuracy on test set. Meanwhile, some other classical network structures were compared as baselines. After forward inference process of the neural network model, a logic enhancement algorithm was raised and applied in this paper, due to the prediction error between two classes. Experiments were conducted on real-time videos, achieving satisfying performance. © (2019) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only. |
Persistent Identifier | http://hdl.handle.net/10722/278007 |
ISSN | 2023 SCImago Journal Rankings: 0.152 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Zhao, Y | - |
dc.contributor.author | Fok, WWT | - |
dc.contributor.author | Chan, CW | - |
dc.date.accessioned | 2019-10-04T08:05:38Z | - |
dc.date.available | 2019-10-04T08:05:38Z | - |
dc.date.issued | 2019 | - |
dc.identifier.citation | 2nd International conference on Image, Video Processing and Artificial Intelligence (IVPAI2019), Shanghai, China, 23-25 August 2019. In Proceedings of SPIE, 2019, v. 11321, p. 113212N:1-8 | - |
dc.identifier.issn | 0277-786X | - |
dc.identifier.uri | http://hdl.handle.net/10722/278007 | - |
dc.description.abstract | In recent years, human action analysis is a focal point in video processing, especially on action recognition and safety surveillance. It always performs as an auxiliary tool to minimize the manpower-resource on special tasks. This paper explores the human action analysis in a specified situation, based on the human posture extraction by pose-estimation algorithm. Deep neural network (DNN) methods was used, composed of residual learning blocks for feature extraction and recurrent neural network for time-series data learning. All these modules can be applied on real-time videos, classifying different security levels of actions between two people, with 91.8% accuracy on test set. Meanwhile, some other classical network structures were compared as baselines. After forward inference process of the neural network model, a logic enhancement algorithm was raised and applied in this paper, due to the prediction error between two classes. Experiments were conducted on real-time videos, achieving satisfying performance. © (2019) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only. | - |
dc.language | eng | - |
dc.publisher | SPIE - International Society for Optical Engineering. The Journal's web site is located at http://spie.org/x1848.xml?WT.svl=mddp2 | - |
dc.relation.ispartof | 2nd International conference on Image, Video Processing and Artificial Intelligence (IVPAI2019) | - |
dc.relation.ispartof | SPIE - International Society for Optical Engineering. Proceedings | - |
dc.rights | SPIE - International Society for Optical Engineering. Proceedings. Copyright © SPIE - International Society for Optical Engineering. | - |
dc.subject | Human action analysis | - |
dc.subject | Long-Short-Term-Memory | - |
dc.subject | Neural network | - |
dc.subject | Pose estimation | - |
dc.subject | Residual learning | - |
dc.subject | Video processing | - |
dc.subject | Violence detection | - |
dc.title | Video-based Violence Detection by Human Action Analysis with Neural Network | - |
dc.type | Conference_Paper | - |
dc.identifier.email | Fok, WWT: wilton@hkucc.hku.hk | - |
dc.identifier.authority | Fok, WWT=rp00116 | - |
dc.identifier.doi | 10.1117/12.2543564 | - |
dc.identifier.scopus | eid_2-s2.0-85077816141 | - |
dc.identifier.hkuros | 306899 | - |
dc.identifier.volume | 11321 | - |
dc.identifier.spage | 113212N-1 | - |
dc.identifier.epage | 113212N-8 | - |
dc.identifier.isi | WOS:000511402300093 | - |
dc.publisher.place | United States | - |
dc.identifier.issnl | 0277-786X | - |