Anomalous Event Detection in Surveillance Videos using Spatio-temporal Autoencoders
Journal: International Journal of Advanced Trends in Computer Science and Engineering (IJATCSE) (Vol.11, No. 3)Publication Date: 2022-06-12
Authors : Sai Kiran Singamaneni Sujith Thota Amutha Prabakaran M;
Page : 77-79
Keywords : Computer vision; Deep Learning; Spatiotemporal autoencoder; decoder; convolutional networks; Long short term memory(LSTM).;
Abstract
Surveillance cameras are proliferating, and millions of devices are being used to capture endless footage of surveillance videos. With the advancements in computer vision and deep learning, we can now contemplate these videos to detect anomalies. In this paper, we propose to identify anomalies by using Spatiotemporal autoencoders. The autoencoders based on 3-D convolutional networks will identify anomalies in surveillance footage based on spatiotemporal features. Our architecture consists of two components, an encoder for spatial feature extraction, and a decoder for the reconstruction of frames. Then, abnormal events are identified based on reconstruction loss. We used the Avenue dataset and UCSD dataset for training and evaluation.
Other Latest Articles
- Age Determination of Harbour Porpoise (Phocoena phocoena relicta) from the Bulgarian BlackSeaCoast
- Pluteus fenzlii (Pluteaceae, Agaricales) RediscoveredintheBalkan Peninsula after over 150 Years Gap
- First Record of Heteropterus morpheus (Pallas, 1771) (Lepidoptera: Hesperiidae) from the Republic of
- The concept, signs and place of executive power in the mechanism of the modern state
- Features of the Komi Republic budget
Last modified: 2022-06-19 20:05:53