Inspiration
Enforcing government regulations for masks is a challenging task for essential workers. MaskerAId has the potential to help identify people that are not wearing a mask and the system can nudge or remind these people to wear a mask.
What it does
Detects if you are wearing a mask through video or image data.
How we built it
Built an image classifier with Keras and Tensorflow and integrated it into a web app with Streamlit.
Challenges we ran into
- Deciding on an initial problem definition that was feasible given our skills and the timeframe
- Finding suitable datasets
Accomplishments that we're proud of
- Creating our first deep learning models for some of us
- Doing our first hackathon
- Creating a really cool and useful app
What we learned
- Learning how to handle video data in python
- Learning streamlit and how it can produce simple and easy web-apps
What's next for MaskerAId!
- Bug/ performance fixes
- Multi-face recognition
- Testing it in a real environment such as college accommodation or at the entrance of a grocery store
- Improvements in the model accuracy
- Web deployment (Streamlit URL sharing)
Log in or sign up for Devpost to join the conversation.