Security surveillance is critical to social harmony and people’s peaceful life. It has a great impact on strengthen- ing social stability and life safeguarding. Detecting anoma- ly timely, effectively and efficiently in video surveillance remains challenging. This paper proposes a new approach, called S 2 -VAE, for anomaly detection from video data. The S2-VAE consists of two proposed neural networks: a Stacked Fully Connected Variational AutoEncoder (SF - VAE) and a Skip Convolutional VAE (SC-VAE). The SF -VAE is a shallow generative network to obtain a Gaussian mixture like model to fit the distribution of the actual data. The SC -VAE, as a key component of S 2 - VAE, is a deep generative network to take advantages of CNN, VAE and skip connections. Both SF -VAE and SC-VAE are efficient and effective generative networks and they can achieve better performance for detecting both local abnormal events and global abnormal events. The proposed S 2 -VAE is evaluated using four public datasets. The experimental results show that the S2-VAE outperforms the state-of-the-art algorithms.
|Number of pages||10|
|Journal||IEEE Transactions on Information Forensics and Security|
|Publication status||Accepted/In press - 13 Oct 2018|
- anomaly detection,
- anomaly detection, Variational AutoEncoder
- loss function
Wang, T., Qiao, M., Lin, Z., Li, C., Snoussi, H., Liu, Z., & Choi, C. (Accepted/In press). Generative Neural Networks for Anomaly Detection in Crowded Scenes. IEEE Transactions on Information Forensics and Security, 0(0), 1-10. .