For this exercise, I made a “Girlboss identifier” machine. A corresponding LED would light up in response to whether “gatekeep”, “gaslight” or “girlboss” was said into in the microphone.

I first trained Google’s Teachable Machine with sound samples for each word as well as background noise. I then let Teachable Machine do the work. Here’s the resulting model:

https://teachablemachine.withgoogle.com/models/iWxErI2RF/

I used adjusted sample code provided by my professor to take in the three inputs in p5.js and to connect to the correct port on my computer.

[P5.js sketch](https://editor.p5js.org/marthajanicki/sketches/VCPNTMgSs])

I also adjust code provided by my professor for the Arduino to output settings on three LEDs instead of 2.

https://gist.github.com/marthajanicki/19aff5889ba636cf8fb1f260349d1767

Here is the output as a video:

IMG_3079_480p.mov

Alternative ideas:

• Play doorbell when door opens or closes.
• Turn motor left or right based on voice commands.
• Sound output says boo when a mask goes on the face. Or a game like Celebrity where when a mask goes on the face, sound output says: “Who are you?” then runs through a list of random names. The person with the mask waits until they hear the name of their character then takes off the mask.