inspiration
The terrible consequences of drunk driving and the need to address this issue in our society served as the impetus for SafeRouteGuardian. Roughly 1,566,622 people lost their lives in drunk driving in 2022.
What it does
To assess a person's state of sobriety, SafeRouteGuardian is a solution that combines audio and facial recognition technology. As a result, we are able to determine whether they are competent to drive.
How we built it
The facial recognition algorithm of SAfeRouteGuardian is based on a specially designed neural network that has been taught to identify indicators of intoxication, like dilated pupils, shakiness, and an overall messy appearance. By transcribing spoken words and comparing them to the original phrase, its audio recognition system can detect differences in enunciation, intonation, or pronunciation that may be signs of potential intoxication.
Challenges we ran into
One of the biggest issues we had was with the data. We had to gather our own data because we had to build the networks from scratch, and not much of it was available online. Nevertheless, we managed to locate a few elusive data points and put together a training set.
We also struggled to connect our networks to our website, but after many sleepless hours, we succeeded!
Accomplishments that we're proud of
Users may easily use SafeRouteGuardian on the go, anywhere at any time, thanks to its mobile-responsive design, accessibility, and convenience.
What we learned
We gained a ton of knowledge about theory and application in the field of neural networks during the creation process. In addition, a plethora of algorithms, including the Levenshtein distance algorithm, were covered.
What's next for SafeRouteGuardian
SafeRouteGuardian aims to prevent drunk driving occurrences in order to change the world and save lives. Our goal is to lower the number of drunk driving collisions and save lives—not only in India but all over the world. We want to use user input and a mobile app to do this
Log in or sign up for Devpost to join the conversation.