I’m working on some ideas using interactive machine learning with the morph. I just added an example to my openFrameworks addon, ofxRapidLib, here on github.
Currently, I’m testing on OS 10.12 and ofx 0.9.8. I haven’t been brave enough to jump to ofx 0.10 yet.
This demo identifies different objects resting on the morph and plays a different audio loop for each. The classification is sensitive to position, so it helps to record examples of objects in multiple locations on the surface. I’m having good luck with coffee cups and bottles.