A Comparative Study of Aggressive Driving Behavior Recognition Algorithms Based on Vehicle Motion Data

Abstract

Aggressive driving, amongst inappropriate driving behaviors, is largely responsible for leading to traffic accidents, which threatens both the safety and property of human beings. With the objective to reduce traffic accidents and improve road safety, effective and reliable aggressive driving recognition methods, which enables the development of driving behavior analysis and early warning systems, are urgently needed. Most recently, the research focus of aggressive recognition has shifted to the use of vehicle motion data, which has emerged as a new tool for traffic phenomenon explanation. As aggressive driving corresponds to sudden variations in data, they can be recognized based on the recorded vehicle motion data. In this paper, several kinds of anomaly recognition algorithms are studied and compared, using the motion data collected by the accelerometer and gyroscope of a smartphone mounted on the vehicle. Gaussian mixture model (GMM), partial least squares regression (PLSR), wavelet transformation, and support vector regression (SVR) are considered as the representative algorithms of statistical regression, time series analysis, and machine learning, respectively. These algorithms are evaluated by the three widely used validation metrics, including F 1 -score, precision, and recall. The empirical results show that GMM, PLSR, and SVR are promising methods for aggressive driving recognition. GMM and SVR outperform PLSR when only single-source dataset is used. The improvement of F 1 -score is almost 0.1. PLSR performs the best when multi-source datasets are used, and the F 1 -score is 0.77. GMM and SVR are more robust to hyperparameter. In addition, incorporating multi-source datasets helps improve the accuracy of aggressive driving behavior recognition. INDEX TERMSAggressive driving recognition, Gaussian mixture model, partial least squares regression, wavelet transformation, support vector regression, vehicle motion data. YONGFENG MA received the Ph.D. degree in transportation engineering from Southeast University, Nanjing, China, in 2007. He is currently an Associate Professor with the School of Transportation, Southeast University. He has been a Visiting Scholar with the Center for Road Safety, School of Civil Engineering, Purdue University, USA, from 2014 to 2015. He has authored or co-authored over 40 articles in international journals and conferences. His research interests include intelligent transportation systems, traffic safety, driving behavior analysis, and data mining.

Secure Accelerometer-Based Pairing of Mobile Devices in Multi-Modal Transport

“…In trusted environments, features from accelerometers can be used to determine if two devices are held by the same person [20], to detect driving patterns [35] or abnormal driving behaviour [22], [41] or to recognize human activities [30]. Distinguishing between different transportation modes based on accelerometer data, e.g., car, bike, bus, etc., has also gained momentum [10], [18], [29], [31], [40].…”

Wide or Narrow? A Visual Attention Inspired Model for View-Type Classification

“…These metrics have been applied in many studies to evaluate the performance of classification algorithm [55]- [57]. All of them differentiate the correct classification of labels within different classes [58], [59], which are defined as follows,…”

Sensor-Based Extraction Approaches of In-Vehicle Information for Driver Behavior Analysis

“…Some studies also utilize wearable sensors to monitor the driver’s condition [ 16 ]. As an alternative to installing additional sensors in vehicles, some researchers have utilized conventional smartphones as sensing devices [ 17 , 18 , 19 , 20 , 21 ]. The use of a smartphone as a sensor is inexpensive, but the collected information is limited to the functions supported by the smartphone.…”