Anomaly Detection For Autonomous Driving Public Transports

Anomaly Detection is critical for the public safety, in autonomous driving public transports. An anomaly detection model for autonomous driving public transports is proposed in this paper, which is based on Variational Autoencoder (VAE). In the model, a fusion feature is adopted as the feature expression of the input. In the VAE, the convolutional layers and deconvolutional layers are used to process the input images. And the proposed model is evaluated on a dataset. The results show that it is effective for some specific kinds of anomalies.