Abstract :

The In-vehicle Environment Was Found To Generate A Diverse Range Of Acoustic Signals Originating From Driver Activities, Engine Operations, Road Interactions, And Surrounding Traffic Conditions. These Audio Events Carry Significant Information Related To Driver Behaviour And Vehicle Safety. With The Advancement Of Intelligent Transportation Systems, The Need For Automated And Real-time Monitoring Of Driver Alertness Using In-vehicle Audio Has Gained Considerable Importance. Traditional Approaches Relied On Manual Observation Or Basic Signal Processing Techniques, Which Were Limited In Handling Complex Acoustic Patterns And Large-scale Audio Data Generated In Modern Vehicles. To Overcome These Limitations, A Deep Learning-based In-vehicle Audio Event Detection Framework Was Developed For Enhancing Driver Alertness Monitoring And Safety. The Proposed System Utilized The Waveform Language Model (WavLM), A Transformer-based Architecture, To Extract Deep Acoustic Feature Representations From Raw Audio Signals. These Features Captured Both Temporal And Contextual Information, Enabling Effective Modelling Of Complex In-vehicle Sound Patterns. The Extracted Feature Vectors Were Used To Train Multiple Machine Learning Classifiers, Including Categorical Boosting Classifier (CBC), Histogram Gradient Boosting Classifier(HGBC) , Extra Trees Classifier (ETC), And A Proposed Tree-Based Generalized Additive Model (TGAM). The System Was Designed To Perform Both Main-class And Sub-class Audio Event Classification To Accurately Identify Driver-related Events Such As Distractions, Alerts, And Environmental Sounds. The Performance Of The Models Was Evaluated Using Standard Metrics Such As Accuracy, Precision, Recall, And F1-score. Experimental Results Demonstrated That The Proposed TGAM Model Achieved Superior Performance, With An Accuracy Of 99.92% For Mainclass Classification And 99.58% For Sub-class Classification. The Integration Of Deep Learning-based Feature Extraction With Advanced Classification Techniques Significantly Improved The Reliability And Efficiency Of In-vehicle Audio Event Detection Systems, Thereby Contributing To Enhanced Driver Alertness Monitoring And Overall Road Safety.

Published:

22-6-2026

Issue:

Vol. 26 No. 6 (2026)

Page Nos:

1309-1318

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite