Really impressed by these generated voice and piano samples: https://deepmind.com/blog/wavenet-generative-model-raw-audio/ … --waiting for entire raw audio music tracks next!
-
We need a Keras implementation of Wavenet!
it's a residual stack of dilated convs. Should be very easy to implement in Keras with functional API + AtrousConv2D layer.
-
-
you should have a go at it! Maybe train it on the million song DS. If you could generate 10s music samples that would be huge.
-
Yes, I'm planning on doing something like that! Now that I'm done with my summer internship, I have more time for these things :)
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.