Authors

Inspiration

Coronavirus or not, the most common mistake people make in their day-to-day lives is touching their face. This is an intuitive habit, but one that gets bacteria spreading rapidly. People touch their face when they are working, when they are bored or tired, or simply whenever they are irritated, and it is a difficult task to control.

What it does

We developed 'HandsOff' , a web application that alerts the user when they are touching their face.

How we built it

The application's hand detection system function based on a machine learning model of hands on faces and hands in general. HandsOff makes use of the user's computer webcam, and when the user touches his/ her face, plays an audio to remind the user of the deadly pandemic.

Challenges we ran into

The biggest challenge we faced was creating the correct Model that would be able to detect hands over face with a high accuracy score. After multiple trials and error the current accuracy score is 0.79.

Accomplishments that we're proud of

We are proud we were able to create our first Machine Learning based Web Application that is able to detect hand presence on the face.

What we learned

We learned a lot of essential computer science skills through this opportunity and hope our application provides users with a simple yet convenient solution that helps them stay safe from the ongoing pandemic.

What's next for HandsOff

Next for HandsOff is developing the application on different kinds of devices such as Android and iOS mobiles.

Try It out

Hackathons

Technologies

css, handtrack.js, html, javascript, machine-learning

Devpost Software Identifier

260476