I made these 2 demo videos for the plenary as appendices the the main presentation which will become public in due time. Since these 2 videos are long, slow, and build from code that is already shared in the distro, I thought I would post them here, so people could feedback on how they are useful or not. They are under-edited to be thorough, but maybe they are not enough, or too much, or whatever. Comments welcome.
The first is building from the chaotic synth example (8b, this is the SC version but the Max one is exactly the same) and then it uses MFCC to try to control it via regression of preset-to-audio-description.
The second is building on my pitch regression example, and shows how KNN and MLP each behave differently to the task I don’t want to lead you too much but maybe there is something useful in there.
I hope some of you enjoy some of it.