Road User Abnormal Trajectory Detection using a Deep Autoencoder
This work addresses a domain-specific problem for traffic safety and monitoring, but it is incremental as it applies an existing method (deep autoencoder) to a new dataset with data augmentation.
The paper tackles the problem of detecting abnormal trajectories of road users at traffic intersections by proposing a deep autoencoder trained on augmented normal data, achieving better performance than classical outlier detection methods on four outdoor urban scenes.
In this paper, we focus on the development of a method that detects abnormal trajectories of road users at traffic intersections. The main difficulty with this is the fact that there are very few abnormal data and the normal ones are insufficient for the training of any kinds of machine learning model. To tackle these problems, we proposed the solution of using a deep autoencoder network trained solely through augmented data considered as normal. By generating artificial abnormal trajectories, our method is tested on four different outdoor urban users scenes and performs better compared to some classical outlier detection methods.