Have you ever wanted to cross the street but found a horde of speeding cars in front of you? We often find ourselves in this situation when we aren't sure if a gap between the cars is large enough for us to safely cross the street. Crossy Road, a piece of novel headband technology, aims to conquer this problem.
A lot of times we want to walk across the street, but we need to wait too long. Also, 3000 pound metal cars are speeding down the street in front of us. Clearly there is something to be improved upon here. We need a faster and most importantly safer way to cross the street.
We built a way to play crossy road in real life with a hack. You just run in the middle but it's okay because you have two lives and this is just a tutorial.
What it does
Crossy road is a novel sensory device built off of two cameras, one on each side of the head. By utilizing an object recognition server we calculate the velocity of the cars to predict the position of cars in the future, allowing the user to know whether it is safe to cross the street or not.
How we built it
The server is made with python. The app is made with swift. The device is made with extremely high quality deluxe (TM) state of the art corrugated cardboard platinum PLUS edition.
Challenges we ran into
Getting the object detection model to run with high accuracy was hard.
Accomplishments that we're proud of
It works!!!!!! I was able to cross the street.