Vehicle classification from low-frequency GPS data with recurrent neural networks. (June 2018)

• Record Type: Journal Article
• Authors: Simoncini, Matteo; Taccari, Leonardo; Sambo, Francesco; Bravi, Luca; Salti, Samuele; Lori, Alessandro
• Is Part Of: Transportation research. Volume 91(2018)
• Page Start: 176
• Page End: 191
• Publication Date: 2018-06
• Subjects: Vehicle classification -- GPS -- Sequence classification -- Recurrent neural networks; Transportation -- Periodicals; Transportation -- Technological innovations -- Periodicals; 388.011
• DOI: 10.1016/j.trc.2018.03.024
• Languages: English
• ISSNs: 0968-090X

Abstract

Highlights: Vast amount of GPS data from connected vehicles are collected every minute. We study the effectiveness of recurrent neural networks for vehicle classification from GPS data. An end-to-end classification system is applied to a binary and a multi-class problem. LSTM-based recurrent neural networks outperform the existing state-of- the-art. Detailed experimental analysis and insights on a very large GPS dataset are provided. Abstract: The categorization of the type of vehicles on a road network is typically achieved using external sensors, like weight sensors, or from images captured by surveillance cameras. In this paper, we leverage the nowadays widespread adoption of Global Positioning System (GPS) trackers and investigate the use of sequences of GPS points to recognize the type of vehicle producing them (namely, small-duty, medium-duty and heavy-duty vehicles). The few works which already exploited GPS data for vehicle classification rely on hand-crafted features and traditional machine learning algorithms like Support Vector Machines. In this work, we study how performance can be improved by deploying deep learning methods, which are recently achieving state of the art results in the classification of signals from various domains. In particular, we propose an approach based on Long Short-Term Memory (LSTM) recurrent neural networks that are able to learn effective hierarchical and stateful representations for temporal sequences. We provide several insights on what the network learns when trained with GPS data and contextual information, and report experiments on a very large dataset of GPS tracks, where we show how the proposed model significantly improves upon state-of-the-art results.