Generative Neural Networks for Anomaly Detection in Crowded Scenes
Security surveillance is critical to social harmony and people's peaceful life. It has a great impact on strengthening social stability and life safeguarding. Detecting anomaly 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 S 2 -VAE consists of two proposed neural networks: a Stacked Fully Connected Variational AutoEncoder (S F -VAE) and a Skip Convolutional VAE (S C -VAE). The S F -VAE is a shallow generative network to obtain a model like Gaussian mixture to fit the distribution of the actual data. The S C -VAE, as a key component of S 2 -VAE, is a deep generative network to take advantages of CNN, VAE and skip connections. Both S F -VAE and S C -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 S 2 -VAE outperforms the state-of-the-art algorithms. The code is available publicly at https://github.com/tianwangbuaa/.
PDFDatasets
Results from the Paper
Task | Dataset | Model | Metric Name | Metric Value | Global Rank | Benchmark |
---|---|---|---|---|---|---|
Abnormal Event Detection In Video | UBI-Fights | s2-VAE | AUC | 0.610 | # 3 | |
Decidability | 0.323 | # 3 | ||||
EER | 0.427 | # 3 | ||||
Semi-supervised Anomaly Detection | UBI-Fights | s2-VAE | AUC | 0.540 | # 5 | |
Decidability | 0.164 | # 5 | ||||
EER | 0.475 | # 5 |