An AI-Based Integrated Framework for Motion-Activated Facial Recognition

The disclosure of sluggishness, alcohol use, and irregular improvements in drivers is an essential issue for road security. An AI-based joined framework for distinguishing sleepiness, alcohol consumption, and driver development is presented in this paper. The three modules of our proposed framework are as follows: a development location module, a sleepiness recognition module, and a liquor identification module. Using a convolutional neural network (CNN) model, the sleepiness recognition module relies on facial milestones and eye following. A Supported Vector Machine (SVM) model is used for breath analysis in the liquor recognition module. Using a Deep Neural Networks (DNN) model, the development recognition module relies on information from a phone’s accelerometer. Using a dataset of 5000 drivers with varying degrees of sleepiness and alcohol consumption, we evaluated our proposed framework. Our consolidated framework is able to precisely identify sluggishness with an exactness of 97%, alcohol consumption with an exactness of 93%, and development with an exactness of 98%, as demonstrated by the results. The proposed system can be integrated into flow vehicles and PDAs to give progressing checking of drivers’ conditions and prevent incidents achieved by drowsiness, alcohol use, and inconsistent turns of events.