MusicNet aims to give machine learning algorithms a taste for Beethoven
We’ve got machine learning systems captioning images, synthesizing speech and translating between languages, but it seems we’ve been neglecting their cultural development. Would you believe the future rulers of this planet are largely ignorant of classical music altogether? It’s bad enough humanity will be enslaved — but by an entity so vulgar? A new project looks to change that with a data set of pieces from great composers, annotated so the computers can follow along.MusicNet, conceived and compiled by University of Washington researchers, is an attempt to create a data set that is to analyzing classical music what the widely used ImageNet is for computer vision. It isn’t the first time machine learning has encountered music, by a long shot, but MusicNet is more about establishing a standard set of…
Link to Full Article: MusicNet aims to give machine learning algorithms a taste for Beethoven
Sharing
