Publication Type : Conference Proceedings
Publisher : Advances in Intelligent Systems and Computing
Source : Advances in Intelligent Systems and Computing, Springer Verlag, Volume 815, p.239-247 (2019)
Url : https://www.scopus.com/inward/record.uri?eid=2-s2.0-85057124299&doi=10.1007%2f978-981-13-1580-0_23&partnerID=40&md5=aceb2180b879f109b8ac76543538c737
ISBN : 9789811315794
Keywords : Anomaly detection, Convolutional neural network, Deep neural networks, Environmental change, Event classification, Feature learning, High dimensionality, Its efficiencies, Learning algorithms, Monitoring, Neural networks, Security systems, Video recording, Video surveillance
Campus : Bengaluru
School : Department of Computer Science and Engineering, School of Engineering
Department : Computer Science
Year : 2019
Abstract : Anomaly detection in video surveillance data is very challenging due to large environmental changes and human movement. Additionally, high dimensionality of video data and video feature representation adds to these challenges. Many machine learning algorithms failed to show accurate results and it is time consuming in many cases. The semi supervised nature of deep learning algorithms aids in learning representations from the video data instead of hand crafting the features for specific scenes. Deep learning is applied to handle complicated anomalies to improve the accuracy of anomaly detection due to its efficiency in feature learning. In this paper, we propose an efficient model to predict anomaly in video surveillance data and the model is optimized by tuning the hyperparameters. © Springer Nature Singapore Pte Ltd. 2019.
Cite this Research Publication : K. Kavikuil and Amudha J., “Leveraging deep learning for anomaly detection in video surveillance”, Advances in Intelligent Systems and Computing, vol. 815. Springer Verlag, pp. 239-247, 2019.