Inspiration
We all are well-versed of this project's purpose
What it does
Determines through your webcam whether you're wearing a mask or not
How I built it
Using python libraries, Keras and the CNN algorithm to preprocess the images and to build and train the model while OpenCV was primarily used to connect the webcam to the model.
Challenges I ran into
The first models helped me understand the CNN algorithm but didn't produce reproducible results. The model was overfitting because of the small number of datasets used which is when I decided to expand the number of datasets which helped improve the model's validation and test data's accuracy.
Accomplishments that I'm proud of
Creating a working project that uses the webcam to help visualize the completion of the classification model in real-time
What I learned
How to train, and re-train models to be able to refine its ability to predict the correct results on test data never before seen by the model.
What's next for Face Mask Detector - MAIS 202 Project
I tried initially connecting this project to a raspberry pi but the os download kept corrupting the sd cards which prevented me from continuing with the idea, but it's definitely something I want to try.
Log in or sign up for Devpost to join the conversation.