Identifying driving safety profiles from smartphone data using unsupervised learning

Mantouka, Eleni G.; Barmpounakis, Emmanouil; Vlahogianni, Eleni I.

doi:10.1016/j.ssci.2019.01.025

Cited by 86 publications

(28 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In other recent work [11] researchers propose an approach utilizing smartphone sensor data (acceleration in m/s2 with an accelerometer, angular velocity in rad/sec with gyroscope, and speed and vehicle position with GPS) recognizing unsafe driving styles based on a two-stage clustering approach and using the information on harsh events occurrence, acceleration profile, mobile usage, and speeding. This approach consists of the following steps: initially, clustering is applied in order to separate aggressive from non-aggressive trips; a second level clustering aided to distinguish normal trips from unsafe trips; thereby, trips were classified into six distinct groups ranked by the importance of driving safety: safe, aggressive, risky (speeding), distracted (mobile usage), aggressive/risky, and aggressive/distracted behavior.…”

Section: A Smartphone-based Driver Monitoringmentioning

confidence: 99%

“…The database contains the number of directories equal to the number of speakers (20). Then, the directory of each speaker includes subdirectories equal to the number of uttered phrases (50 according to our dictionary) as well as metadata file that describes the speaker: speaker (8)(9)(10)(11)(12). Thus, at the bottom of our structure we have audio-visual files of the same speaker uttering the same phrase during different recording conditions.…”

Section: Multimodal Corpus Rusavicmentioning

confidence: 99%

See 1 more Smart Citation

Multimodal Corpus Design for Audio-Visual Speech Recognition in Vehicle Cabin

et al. 2021

View full text Add to dashboard Cite

This paper introduces a new methodology aimed at comfort for the driver in-the-wild multimodal corpus creation for audiovisual speech recognition in driver monitoring systems. The presented methodology is universal and can be used for corpus recording for different languages. We present an analysis of speech recognition systems and voice interfaces for driver monitoring systems based on the analysis of both audio and video data. Multimodal speech recognition allows using audio data when video data are useless (e.g. at nighttime), as well as applying video data in acoustically noisy conditions (e.g., at highways). Our methodology identifies the main steps and requirements for multimodal corpus designing, including the development of a new framework for audiovisual corpus creation. We identify the main research questions related to the speech corpus creation task and discuss them in detail in this paper. We also consider some main cases of usage that require speech recognition in a vehicle cabin for interaction with a driver monitoring system. We also consider other important use cases when the system detects dangerous states of driver's drowsiness and starts a question-answer game to prevent dangerous situations. At the end based on the proposed methodology, we developed a mobile application that allows us to record a corpus for the Russian language. We created RUSAVIC corpus using the developed mobile application that at the moment a unique audiovisual corpus for the Russian language that is recorded in-the-wild condition. INDEX TERMS Driver monitoring, automatic speech recognition, multimodal corpus, human-computer interaction.

show abstract

Section: A Smartphone-based Driver Monitoringmentioning

confidence: 99%

Section: Multimodal Corpus Rusavicmentioning

confidence: 99%

Multimodal Corpus Design for Audio-Visual Speech Recognition in Vehicle Cabin

et al. 2021

View full text Add to dashboard Cite

show abstract

“…On the contrary, identifying risky or accident-prone drivers could facilitate more effective traffic safety work, coupled with the most appropriate incentives to avoid them. It is vital for the achievement of safe and efficient driving [16,17]. The authors of [18] suggest a framework for identifying driver behavior that takes three anomalies in accelerations and directions into account, which are lane changes, excessive speed, and abrupt movements, while [19] proposed a data collection process using a set of vehicle sensors to identify a series of driving maneuvers to categorize driver features or assess the driver's skills.…”

Section: Related Workmentioning

confidence: 99%

Identification of Driving Safety Profiles in Vehicle to Vehicle Communication System Based on Vehicle OBD Information

et al. 2021

View full text Add to dashboard Cite

Driver behavior is a determining factor in more than 90% of road accidents. Previous research regarding the relationship between speeding behavior and crashes suggests that drivers who engage in frequent and extreme speeding behavior are overinvolved in crashes. Consequently, there is a significant benefit in identifying drivers who engage in unsafe driving practices to enhance road safety. The proposed method uses continuously logged driving data to collect vehicle operation information, including vehicle speed, engine revolutions per minute (RPM), throttle position, and calculated engine load via the on-board diagnostics (OBD) interface. Then the proposed method makes use of severity stratification of acceleration to create a driving behavior classification model to determine whether the current driving behavior belongs to safe driving or not. The safe driving behavior is characterized by an acceleration value that ranges from about ±2 m/s2. The risk of collision starts from ±4 m/s2, which represents in this study the aggressive drivers. By measuring the in-vehicle accelerations, it is possible to categorize the driving behavior into four main classes based on real-time experiments: safe drivers, normal, aggressive, and dangerous drivers. Subsequently, the driver’s characteristics derived from the driver model are embedded into the advanced driver assistance systems. When the vehicle is in a risk situation, the system based on nRF24L01 + power amplifier/low noise amplifier PA/LNA, global positioning system GPS, and OBD-II passes a signal to the driver using a dedicated liquid-crystal display LCD and light signal. Experimental results show the correctness of the proposed driving behavior analysis method can achieve an average of 90% accuracy rate in various driving scenarios.

show abstract

“…Speeding, Tailgating, driver's association with crash or near-crash were used to identify the aggressive drivers and results showed that aggressive drivers had significantly higher values of the two jerk-based metrics. Most recently the authors in [71] presented a two-stage clustering approach for the detection of unsafe driving styles by utilizing driving data and informa-tion on mobile usage, harsh events occurrence, speeding and acceleration profile with increasing importance with respect to safety. In this way, trips have been categorized into six distinct groups (Aggressive trips include Aggressive trips, Distracted trips and Risky trips).…”

Section: B Human Driver Aggressive Driving Behavior (Hadb)mentioning

confidence: 99%

Detecting Human Driver Inattentive and Aggressive Driving Behavior Using Deep Learning: Recent Advances, Requirements and Open Challenges

2020

View full text Add to dashboard Cite

Human drivers have different driving styles, experiences, and emotions due to unique driving characteristics, exhibiting their own driving behaviors and habits. Various research efforts have approached the problem of detecting abnormal human driver behavior with the aid of capturing and analyzing the face of driver and vehicle dynamics via image and video processing but the traditional methods are not capable of capturing complex temporal features of driving behaviors. However, with the advent of deep learning algorithms, a significant amount of research has also been conducted to predict and analyze driver's behavior or action related information using neural network algorithms. In this paper, we contribute to first classify and discuss Human Driver Inattentive Driving Behavior (HIDB) into two major categories, Driver Distraction (DD), Driver Fatigue (DF), or Drowsiness (DFD). Then we discuss the causes and effects of another human risky driving behavior called Aggressive Driving behavior (ADB). Aggressive driving Behavior (ADB) is a broad group of dangerous and aggressive driving styles that lead to severe accidents. Human abnormal driving behaviors DD, DFD, and ADB are affected by various factors including driver experience/inexperience of driving, age, and gender or illness. The study of the effects of these factors that may lead to deterioration in the driving skills and performance of a human driver is out of the scope of this paper. After describing the background of deep learning and its algorithms, we present an in-depth investigation of most recent deep learning-based systems, algorithms, and techniques for the detection of Distraction, Fatigue/Drowsiness, and Aggressiveness of a human driver. We attempt to achieve a comprehensive understanding of HIADB detection by presenting a detailed comparative analysis of all the recent techniques. Moreover, we highlight the fundamental requirements. Finally, we present and discuss some significant and essential open research challenges as future directions.

show abstract

Identifying driving safety profiles from smartphone data using unsupervised learning

Cited by 86 publications

References 23 publications

Multimodal Corpus Design for Audio-Visual Speech Recognition in Vehicle Cabin

Multimodal Corpus Design for Audio-Visual Speech Recognition in Vehicle Cabin

Identification of Driving Safety Profiles in Vehicle to Vehicle Communication System Based on Vehicle OBD Information

Detecting Human Driver Inattentive and Aggressive Driving Behavior Using Deep Learning: Recent Advances, Requirements and Open Challenges

Contact Info

Product

Resources

About