The Web-App detects if the person is wearing a face mask(using a neural network) or not, and can then allow the person inside the building depending on if he/she is wearing a mask or not.
The video explaining the project can be found in the same drive, whose link has been mentioned below in the next point.
You must download the model_weights.h5 file from the drive : https://drive.google.com/drive/folders/16Mf1SiB9X0mBpnUZZwlSM96bxCeKnc3I and place it in the same folder as other files available on github repo before running the the project on your PC.
You can also see the dataset on which model was trained in the same drive folder named as "Dataset".
To run the project, put all the files in github repo and model_weights.h5 in a folder and run the command : "streamlit run ./WebPage.py"
- you might get some error regarding required libraries, make sure you have them installed on the PC(like: tensorflow, cv2, streamlit, numpy, etc.)