Abstract
Driving behavior classification is an essential real-world requirement in different contexts. In traffic safety, avoiding traffic accidents by taking corrective actions against aggressive behaviors is necessary to protect drivers. Similarly, in the automotive insurance industry, distinguishing between driving behaviors is essential to adopt usage-based insurance (UBI) policies. Also, in the ridesharing industry, monitoring and evaluating driving behaviors is critical for risk assessment and service improvement. This research presents a deep learning-based solution for driving behavior classification using an optimized Stacked-LSTM model based on the signals of smartphone embedded sensors generating two different classification models: three-class and binary. Three-class classification distinguishes between normal, drowsy, and aggressive behaviors to support advanced driver-assistance systems (ADAS). Binary classification differentiates between aggressive and non-aggressive behaviors to support commercial applications, such as ridesharing services and automotive insurance services based on UBI. Our time-series classification models have been evaluated on the public UAH-DriveSet dataset. Using the proper number and type of features, the optimum factor of upsampling for the raw signals, and the optimum time-series window size, our proposed Stacked-LSTM model made a breakthrough in the F1-score when applied to the aforementioned dataset. The achieved scores are 99.49% and 99.34% for the Three-class and binary classification models, respectively. Comparisons with state-of-the-art models, our three-class classification model surpassed the highest published F1-score of 91% by 8.49% when applied to the aforementioned dataset.
Highlights
Driving behavior can be defined as the way in which individuals prefer to drive and the subsequent habits they gain over the years [1]; so, drivers’ driving behavior plays a significant role in ensuring road traffic safety
The Long ShortTerm Memory (LSTM) model achieved an F1-score of 92.12% compared to SVM, Hidden Markov Model (HMM), Feed Forward Neural Networks (FFNN), Fusion-recurrent neural network (RNN)-Exp-Loss (F-RNNEL) models that achieved an F1-score of 65.40%, 75.72%, 85.76%, and 87.56%, respectively
The LSTM-FCN model achieved an F1-score of 95.88% compared to RF, Adaboost, and ResNet models that achieved an F1-score of 94.11%, 92.75%, and 88.29%, respectively
Summary
Driving behavior can be defined as the way in which individuals prefer to drive and the subsequent habits they gain over the years [1]; so, drivers’ driving behavior plays a significant role in ensuring road traffic safety. Recent researches have proposed different methodologies for classifying driving behaviors using the signals of smartphone sensors These methodologies can be categorized based on (1) the features that can be extracted and derived from the collected data (e.g., acceleration, deceleration, and brake) [6], [11]; (2) the computational models for classifying driving behaviors [19]; (3) driving behavior outputs (e.g., normal, drowsy, or aggressive) [20], [21]; and (4) the performance metrics that evaluate these models [22], [23].
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.