Inspiration
With the prevalence of the COVID-19 and the impact it has had on our world today we wanted to contribute to the fight against the virus. We also wanted to create a project that will not only detect masks but also support the use of Personal Protective Equipment (PPE).
What it does
Our project aims to classify if the user is wearing or not wearing a face-mask based on input from a photo or a live video. Using a convolutional neural network (CNN) we were able to build a model that can predict 0 - Wearing Mask 1 - Not Wearing Mask
How we built it
Numpy for the pre-processing and handling of our data (images) Matplotlib for the visualization of our data Keras to build and train our Convolutional Neural Network OpenCV for real-time video implementation
Contribution
I contributed specifically to the loading of the data and preprocessing of the images. I also contributed to the development of the convolutional neural network and visualizations of the model. Lastly, I presented the project to a panel of judges along with 2 other team members.
Accomplishments that I'm proud of
I'm proud that our group was able to collaborate and take on a data science problem and carry out the workflow from start to finish. I'm also proud that our group was able to successfully demo our project with an accuracy of about 98% and a learning experience all while securing 2nd place at Western AI's 2021 Project Showcase.
What I learned
I learned how to build and compile a Neural Network as well as when to use specific neural networks, such as us choosing a convolutional neural network because of its ability to classify images. Utilizing the OpenCV library to implement real-time computer vision was definitely a great introduction to the field of computer vision. Lastly, tangible experience of working through the Data Science Workflow.
What's Next?
Next up for this project, we would like to implement in a more practical sense, where the model would be able to capture multiple faces and classify each one.
Log in or sign up for Devpost to join the conversation.