English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Detecting abnormal driving behavior is critical for road traffic safety andthe evaluation of drivers' behavior. With the advancement of machine learning(ML) algorithms and the accumulation of naturalistic driving data, many MLmodels have been adopted for abnormal driving behavior detection. Most existingML-based detectors rely on (fully) supervised ML methods, which requiresubstantial labeled data. However, ground truth labels are not always availablein the real world, and labeling large amounts of data is tedious. Thus, thereis a need to explore unsupervised or semi-supervised methods to make theanomaly detection process more feasible and efficient. To fill this researchgap, this study analyzes large-scale real-world data revealing several abnormaldriving behaviors (e.g., sudden acceleration, rapid lane-changing) and developsa Hierarchical Extreme Learning Machines (HELM) based semi-supervised ML methodusing partly labeled data to accurately detect the identified abnormal drivingbehaviors. Moreover, previous ML-based approaches predominantly utilize basicvehicle motion features (such as velocity and acceleration) to label and detectabnormal driving behaviors, while this study seeks to introduce SurrogateSafety Measures (SSMs) as the input features for ML models to improve thedetection performance. Results from extensive experiments demonstrate theeffectiveness of the proposed semi-supervised ML model with the introduced SSMsserving as important features. The proposed semi-supervised ML methodoutperforms other baseline semi-supervised or unsupervised methods regardingvarious metrics, e.g., delivering the best accuracy at 99.58% and the best F-1measure at 0.9913. The ablation study further highlights the significance ofSSMs for advancing detection performance.

Data-driven Semi-supervised Machine Learning with Surrogate Safety Measures for Abnormal Driving Behavior Detection