Assignment 6
Application of data science to present a solution to a real life problem.
1. Problem
With the exponential rise in covid-19 cases the government has made it mandatory for everyone to wear masks in public spaces. However a lot of people are not adhering to these rules thus risking the lives of a lot of people. When public places like bus stations, metro stations, airports and railway stations open, the authorities would require some sort of technology so as to quickly isolate those not wearing masks. Thus there needs to be a system which quickly detects those people who are not wearing masks in public places. For this assignment I have collaborated with Zulfikar Ali who is more accomplished on the coding front and I was more involved on the problem analysis section of the assignment.
2. Solution
We built a machine learning real time face mask detector using Python, Keras, OpenCV and MobileNet which can easily recognize if the person is wearing mask or not. For demonstration purposes we have used the front camera of the laptop but when the solution is implemented it can be integrated with CCTV cameras thus providing a solution for the above problem. This consists of 2 convolutional layers (Two Convo2D 100@3x3). First, you have to load the dataset from data preprocessing. Then you have to configure the convolutional architecture. Since we have two categories (with mask and without mask) we can use binary_crossentropy. You start training for 20 epoch with model checkpoint.
