Automatic reading task prediction on tablet (PC) under the influence of noise and illumination via optimized LSTM

Abstract

Everyone has long recognized smart phones and other mobile technology's capability to enhance cognitive abilities. The majority of studies on smart phones and tablets while driving have focused on distraction, reaction times, and overall safety. In this study, the effects of light and noise on the tablet reading tasks completed by males and females (18 to 45 years old) are investigated. Initially, the collected data are pre-processed via data cleaning process. Subsequently, the most relevant features like improved correlation, improved entropy and statistical features are retrieved from the pre-processed information. The final prediction with respect to the automatic identification of variations in pupil diameter of the left eye, right eye, and reading ability (count of characters read/second) is made using three mathematical models: optimized Long Short-Term Memory1 (LSTM1), optimized LSTM2 and optimized LSTM3 respectively. The optimized LSTM1, optimized LSTM2 and optimized LSTM3 are trained with the fused features. Further, to improve the prediction efficiency of the three mathematical models, weight of them is fine-tuned employing aunique approach known as Grey Wolf Optimization Appended Cuckoo (GWAC). At last, a comparative assessment is undergone to verify the efficiency of the projected model in terms of various measures.