Face Mask Detection with CNN using OpenCV and Keras

Source

Prajna Bhandary's Face Mask Dataset

Process-Flow

• Training set contained 958 images and Testing set contained 406 images belonging to masked and non-masked class
• CNN Binary Classifier was developed with adam optimizer and total trainable parameters in the model were 26529
• Classifier was trained over 10 epochs
• Training Accuracy of 0.983 and Validation Accuracy of 0.980 was achieved

Demo of the Classifier