Inspiration
The current pandemic impacted our society as a whole, to motivate people to not only wear a mask but wear a mask properly, we decided to learn about how AI can be used to detector worn mask.
What it does
The AI can detect if a mask is worn and if it's not worn properly, it would be considered as "not wearing a mask" as wearing a mask improperly is just as dangerous as not wearing one.
How we built it
We got the open-source code of a face mask detector program. By generating our own dataset and training the AI, improperly worn masks would not be considered worn masks.
Challenges we ran into
Jetlag, Hardware, and skill limitations.
Accomplishments that we're proud of
We successfully improved the original AI so that it wouldn't conder improperly wore masks as wearing masks.
What we learned
We learned to basics of AI and image processing. We also learned how to train our AI by generating our own dataset.
What's next for MASK AI
Sometimes, when the mask covered up over 50% of your face, it won't determine whether if the mask is worn because it won't recognize it as a face. However, because the face recognition program is pre-built, it would be difficult for us to train the face recognition program to identify faces cover by masks since it would require us to build our own face recognition program from scratch. Hopefully if the future we can resolve this problem, as our skills improve.
Built With
- cnn
- python
- tensorflow

Log in or sign up for Devpost to join the conversation.