Inspiration
-We would like to protect the community by promoting mask-wearing.
What it does
-The application displays a camera video feed and a banner showing if everyone in the video is wearing a mask.
How we built it
-Trained a Google AutoML model using 580 images of people with and without masks. -Used Google Vision API to detect faces from the camera image. -We then classified the cropped face images using our ML model. -The application is built on React
Challenges we ran into
-Using the API key correctly for the Google Cloud APIs we used, as we could only save it locally.
Accomplishments that we're proud of
-We resized cropped images to improve the performance of the classifier model. -We made our application responsive.
What we learned
-Google Cloud APIs are very impressive. -There is a genius in this team.
What's next for Mask Detector
-Detect multiple people simultaneously. -Prettier design. -Connect our software to electronic entrance gateways to prevent people without masks from entering. -Deploying.
Built With
- google-automl
- google-cloud
- google-vertex-ai
- google-vision
- javascript
- kaggle-datasets
- react
Log in or sign up for Devpost to join the conversation.