Outlier Vehicle Trajectory Detection Using Deep Autoencoders in Santiago, Chile

Abstract

In the last decade, a large amount of data from vehicle location sensors has been generated due to the massification of GPS systems to track them. This is because these sensors usually include multiple variables such as position, speed, angular position of the vehicle, etc., and, furthermore, they are also usually recorded in very short time intervals. On the other hand, routes are often generated so that they do not correspond to reality, due to artifacts such as buildings, bridges, or sensor failures and where, due to the large amount of data, visual analysis of human expert is unable to detect genuinely anomalous routes. The presence of such abnormalities can lead to faulty sensors being detected which may allow sensor replacement to reliably track the vehicle. However, given the reliability of the available sensors, there are very few examples of such anomalies, which can make it difficult to apply supervised learning techniques. In this work we propose the use of unsupervised deep neural network models based on stacked autoencoders to detect anomalous routes in vehicles within Santiago de Chile. The results show that the proposed model is capable of effectively detecting anomalous paths in real data considering validation given by an expert user, reaching a performance of 82.1% on average. As future work, we propose to incorporate the use of Long Short-Term Memory (LSTM) and attention-based networks in order to improve the detection of anomalous trajectories.

Vehicle Trajectory Reconstruction Using Lagrange-Interpolation-Based Framework

“…Based on the difference between the input and the output of the trained stacked autoencoder, the Grubbs criterion and the PauTa criterion are used to evaluate whether data points are outliers [23]. Peralta et al proposed unsupervised deep neural network models based on stacked autoencoders to detect the outliers among position, speed and angular position [24]. Zhao et al proposed a three-step vehicle trajectory reconstruction method with the wavelet transform, Gaussian kernel and Savitzky-Golay filter to reconstruct the trajectory data from 15 intersections.…”

“…Vehicle trajectory data are very important to analyze the microscopic phenomena in transportation systems. GPS devices have usually been used to collect trajectory data in the last decades, but vehicle location sensors can generate a large amount of outlier and observation noise due to a relatively low sampling rate and location error [22,24,31,36]. With the development of multimedia equipment, many transportation researchers have used high-resolution images to extract vehicle trajectory data, and the high-resolution images are from videos and cameras, which are installed on an unmanned aerial vehicle, a moving car or transportation infrastructures.…”