🤖 AI Summary
Addressing the dual challenges of scarce labeled data and inadequate feature representation for detecting anomalous driving behaviors (e.g., abrupt acceleration, frequent lane changes) in real-world scenarios, this paper proposes a semi-supervised learning framework leveraging partially labeled data. We innovatively introduce Event-level Surrogate Measures of Safety (SMoS) — the first systematic integration of such safety-oriented proxy metrics into feature engineering for anomalous driving detection. Furthermore, we design a Hierarchical Extreme Learning Machine (HELM)-based semi-supervised architecture that significantly reduces reliance on large-scale annotated datasets. Evaluated on a real-world naturalistic driving dataset, our model achieves 99.58% accuracy and an F1-score of 0.9913, outperforming state-of-the-art unsupervised and semi-supervised baselines. Ablation studies confirm SMoS contributes a +4.2% F1 improvement, underscoring its critical role in enhancing discriminative feature learning.
📝 Abstract
Detecting abnormal driving behavior is critical for road traffic safety and the evaluation of drivers' behavior. With the advancement of machine learning (ML) algorithms and the accumulation of naturalistic driving data, many ML models have been adopted for abnormal driving behavior detection (also referred to in this paper as anomalies). Most existing ML-based detectors rely on (fully) supervised ML methods, which require substantial labeled data. However, ground truth labels are not always available in the real world, and labeling large amounts of data is tedious. Thus, there is a need to explore unsupervised or semi-supervised methods to make the anomaly detection process more feasible and efficient. To fill this research gap, this study analyzes large-scale real-world data revealing several abnormal driving behaviors (e.g., sudden acceleration, rapid lane-changing) and develops a Hierarchical Extreme Learning Machines (HELM) based semi-supervised ML method using partly labeled data to accurately detect the identified abnormal driving behaviors. Moreover, previous ML-based approaches predominantly utilized basic vehicle motion features (such as velocity and acceleration) to label and detect abnormal driving behaviors, while this study seeks to introduce Surrogate Measures of Safety (SMoS) as input features for ML models to improve the detection performance. Results from extensive experiments demonstrate the effectiveness of the proposed semi-supervised ML model with the introduced SMoS serving as important features. The proposed semi-supervised ML method outperforms other baseline semi-supervised or unsupervised methods regarding various metrics, e.g., delivering the best accuracy at 99.58% and the best F-1 measure at 0.9913. The ablation study further highlights the significance of SMoS for advancing the detection performance of abnormal driving behaviors.
