Drowsiness Detection System

Objective

​

The objective of this project is to implement a drowsiness detection system in vehicles to alert drivers in order to prevent accidents caused by sleep-impaired driving. 

​

Description

​

This project consists of three parts: eye detection, eye classification, and drowsiness classification. Our project emphasizes on stage 2, an inherent image classification problem. Deep learning with CNN would be the most suitable approach because it can extract features, pick up spatial information, and enable weight sharing. It can effectively learn from an eye database and accurately predict eye states.                            

​

My Role

​​

I developed the general architecture of our drowsiness detection model and a baseline model for performance comparison. I also created the facial feature detection and extraction algorithm.

​