Performance Comparison of Deep CNN Models for Detecting Driver’s Distraction

• DOI: 10.32604/cmc.2021.016736
• Publication Date: 2021-05-07T07:22:41Z
• Authors: Kathiravan Sriniv...; Lalit Garg; Debajit Datta; Abdulellah A. Ala...; N. Z. Jhanjhi; Rishav Agarwal; Anmol George Thomas
• Keywords: Neural networks (Computer science) -- Technological innovations; Computer network architectures; 0202 electrical engineering, electronic engineering, information engineering; 02 engineering and technology; Automobile drivers -- Safety measures; Traffic safety -- Technological innovations; Neural networks (Computer science) -- Software

According to various worldwide statistics, most car accidents occur solely due human error. The person driving a needs be alert, especially when travelling through high traffic volumes that permit high-speed transit since slight distraction can cause fatal accident. Even though semiautomated checks, such as speed detecting cameras and barriers, are deployed, controlling errors is an arduous task. key causes of driver’s include drunken driving, conversing with co-passengers, fatigue, operating gadgets while driving. If these distractions accurately predicted, the drivers alerted alarm system. Further, this research develops deep convolutional neural network (deep CNN) models for predicting reason behind distraction. CNN trained using numerous images distracted drivers. performance deepCNNmodels, namely theVGG16,ResNet, Xception network, assessed based on evaluation metrics, precision score, recall/sensitivity F1 specificity score. ResNet model outperformed all other best detection determining drivers’ activities.