Automatic masked face recognition and detection using recurrent neural network compared with linear discriminant analysis to enhance accuracy
Chakali Murali, A. Gayathri · 2025
The objective of the research is to compare the performance of a Recurrent Neural Network method and a Linear Discriminant Analysis technique for automatic face mask detection to find whether a person is wearing a mask or not. In this study, there were two groups with sample size of 100 which were iterated 10 times. The accuracy, performance of the Recurrent Neural Network Classified. Linear Discriminant Analysis and Recurrent Neural Network Analysis are tested by using ClinCalc with a G power of 0.8 with 95% confidence interval. The identification dataset was taken from Kaggle and consists of 20,981 images of 120 face masks. To classify each image using RNN and LDA, 224 rows and 224 columns are used. The outcomes demonstrate that the Novel Recurrent Neural Network has a substantially greater accuracy (94.9%). Recurrent neural networks and Principal Component Analysis shows that it is statistically significant, with (p=0.017) (p<0.05). Linear Discriminant Analysis method is less accurate on average in face mask detection than Novel Recurrent Neural Network.