Application Java mobile Embarquée pour la détection de la somnolence par intelligence artificielle

Abstract

We researched different methods of detecting drowsiness (heart rate, ECG, EEG and others), and ultimately, we chose the ocular method that aims to detect the eyes and report when the eyes close. We have developed a mobile application for ANDROID using the Android Studio environment and the Java programming language. After creating and organizing the folders required by Android studio and then creating and preprocessing our learning base, we implemented the Deep Learning SSD algorithm. We have successfully tested our system on PC. We then integrated our application on an Android smartphone, and obtained results of detection of sleepiness in real time, very conclusive. Our mobile app is functional and efficient.