CovCombat | Devpost

Inspiration - After going to gyms and malls people tend to take off their mask, security can't always be there so automatically detecting whether people are wearing them is crucial.

What it does- It's a program which takes in real time live video feed and whenever it detects a face it is further able to detect whether he/she is wearing a mask or not.

How I built it- it's built on Android java and uses other technologies like google ml kit, tensorflow/keras, openCV.

Challenges I ran into - There were many challenges like where to research and which path to follow luckily my team and I decided to have great connection and decided on particular techs.

To convert the python model into an Android app for that we had to convert the MobileNetV2(.h5 file) to an tensor flow lite file . 2.Getting all those things to work together on Android environment,

Thanks to the internet and stackoverflow we overcame all those.

Accomplishments that I'm proud of- We took on all those problems and worked in a systematic way plus the thing I would call my greatest achievement is meet and making friends with all my teammates.

What I learned- I learned various things ranging from little things like changing certain values for certain variables of a language to combining and linking all the tech to Android java and understanding the basics of ml and deeplearning.

What's next for CovCombat- it shows great promise in today's world whether it be implemented on public transport stations like metros and airport or railway station & specially in closed areas such as gyms, malls and especially it will be an aid to traffic police to identify the person without masks.

Overall we could link it with many industries whether entry without mask shall not be allowed or to learn how the spread of covid is proportional to the number of people not wearing masks .