Digital Eye Strain Detection By Applying Deep Learning Technique To Categories Of Images

Main Article Content

Dr. Deepti Sharma, Dr. Archana B. Saxena, Dr. Deepshikha Aggarwal

Abstract

This work presents the development of an ADES (advanced digital eye strain) detector system. It focuses on digital eye strain detector whose objective is to alert the users of digital devices to alert them if they are using devices for a long period of time. If a user is working on digital devices for prolonged hours, it is necessary that fatigue detection is performed in a non-intrusive way, and that the user should be alert with alarms. Our approach to this open problem uses sequences of images that are 60 s long and are recorded in such a way that the user’s face is visible. To detect whether the user shows symptoms of tiredness or not, a solution is developed focusing on the minimization of false positives. This solution uses a recurrent and convolutional neural network to extract numeric features from images. The accuracy obtained by the proposed system is similar: around 85% accuracy over training data, and 80% accuracy on test data.

Article Details

Section
Articles