Kicker Sound

Inspiration

We already started working on the Infineon Microphone Setup while playing Kicker. A machine learning solution to determine which team scored a goal seemed very unnecessary and overkill. Perfect for a Hackathon

What it does

Idea: Machine learning model determines when a goal was scored. When it's sure a team scored a python script determines the location by comparing the time the sound peak reaches each of the two microphones.

How we built it

Default raspberry and Infineon IM69D130 Microphone configuration

Challenges we ran into

Python machine learning approach did not work because of incompatible library versions. Our initial approach required the ML model to compare stereo input, but Edge Impulse only seems to support mono.

Accomplishments that we're proud of

Raspberry setup, and installation of all required software and libraries got up and working. We managed to solve every problem we ran into even though it was our first headless pi installation.

What we learned

More experience using the Linux Terminal and Python. First time using a ML approach. Working with a Microphone and the Data it provides. ... and many cool Kicker tricks