Safe N Sound

Logo

Inspiration

We wanted to solve the issue of texting and driving. We believe that harnessing the existing technology will make this accessible to everyone, therefore improving road safety.

What it does

The software detects if the driver is not paying attention to the road, and gives them a warning. We have implemented a printed warning for demo purposes, but we will add a loud noise instead.

How we built it

We used OpenCV for facial recognition and Python to implement the learning model.

Challenges we ran into

Initial low accuracy for facial recognition, and low frame rates for real-time analysis.

Accomplishments that we're proud of

Smooth functionality and easy applicability, and a fully functional project on our first hackathon.

What we learned

None of us were familiar with facial recognition software, and we learned how to use new technology and implemented it within 36 hours.

What's next for Safe N Sound

We plan to improve the facial recognition algorithm to recognize more nuanced changes in driver behavior and have different levels of warnings. We plan to implement different audio alerts to improve functionality.

Built With

Submitted to

SD Hacks 2019

Created by

I worked on real time face detection and testing.

Ritwik Sinha
Undergrad with over 5 years of coding experience and 5 months of work experience
I worked on the core logic used for collecting and storing user data for face detection and its implementation in the main recognition program

Samarth Arora
Second Year Computer Science Major interested in Computer Vision, Networking, and Deep Learning
I worked on the face detection and face recognition using openCV.

Palaksh Rungta
Sophomore, Cognitive Sciences w/ Machine Learning
I worked on hardware testing and prototyping. I also helped in the back-end.

Tushita Tandon