Improving Accuracy in Face Mask detection based on Dlib in Raspberry Pi Compared with Viola-Jones method for Pandemic Control

R.Laskhmanan ,D.Sheela
Keywords: Deep Learning, Novel Facemask detection, Raspberry Pi, COVID-19, Dlib, Machine Learning. ,

Abstract

Aim:This project aims to Improve Accuracy in Face Mask detection based on Dlib compared with Viola-Jones method for Pandemic Control. Materials And Methods: Dlib based method and ViolaJones method are chosen as two groups each with 15 samples respectively, which are collected using the training image datasets. G Power =0.8. Result: The independent sample t-test result shows that the Accuracy is improved for Artificial Intelligence-based methods with a mean (96.89) when compared with the Viola-Jones method (73.76) with a two tailed significance (p=0.007). Conclusion:The analysis shows that the accuracy in Dlib based facemask detection is significantly better compared to the Viola-Jones method.