3 ms·
I wonder if a deep fake app for music would be popular. Feed it anything, like the actual Captain Marvel soundtrack, and it generates something similar but dist
by hirundo 8y ago
I wonder if a deep fake app for music would be popular. Feed it anything, like the actual Captain Marvel soundtrack, and it generates something similar but distinctly different. It's at least possible that the quality of the output could be worthwhile background music.
Would the legal implications make this untenable? Could there be an objective measure of the different-ness between the fake and the original that could be used as the legal boundary?
Could generated music based on popular originals become good enough to out compete most human musicians?
If you build it send me a free copy and we'll call it even.
- goldemerald 8y agoI can assure you there is a lot of interest in the Machine Learning field to make deep fakes for music. The problem is deep fakes are not time-series based. Right now a video is treated as a bunch of independent images, whereas you couldn't do that for music/text as there are long term dependencies between time steps. More specifically, there is a recent paper from ICLR about audio GANs (I haven't read it) https://openreview.net/forum?id=H1xQVn09FX https://openreview.net/forum?id=H1xQVn09FX. IMO, this tech should come out (even in app form) in the next 2-4 years.
- hirundo 8y agoThe coming age of infinite mashups. Bruce Springsteen's Fire sung by Elmer Fudd, to infinity. What does a cross between Elvis and John Philip Sousa sound like? Does Jimi Hendrix need a little more Abba? Such an ML generator could explore a space of musical styles with more nodes than there are quarks. Now randomly upload a few to YouTube, let upvotes be the fitness test to select the next generation, and repeat forever. Could this algorithm top the charts or would it devolve into random crap?