Idea & Concept

• The main object of this week is sound classifier, which reminds me of a project for my live coding final from last semester. It is a interactive + performance live game, which audiences could collectively decide the path and choices. More details can be found here if you are interested. :)
• During the performance, when it comes to deciding the path, the method I used back the was really just counting by my eyes. Which is not that efficient since I had to count and then decide the next step. This cause a lot of trouble since I was solo performing, I needed to control the sound, vj, lines, etc. It was just too much. But I feel like with sound classifier. Since would get easier and free me from counting.
• So this project is just a demo for a choosing mechanism, where the audience can shout out their choices and I can just wait until the choice is finalized.
• (Didn’t use arduino since i never officially took class with it… and I lack interest (and confidence) in hardware stuff :/ )

Development

• I began with training Teachable Machine Model. This time it was a easy one since the classification is simple and clear.

  

  And the result is satisfying.

• Then the sketch.

  • I linked the model to p5js.

    

    

    

    

  • Polishing the gotResults function for updating the feedback

    

  • A little add up for visual feedback , and more logic stuff.

  

And here’s the demo!

录屏2025-10-17 下午3.50.59.mov

You can find the code here.