Ollin Boer Bohan

Got it! Not to impose, but do you happen to have an idea for how difficult something like this would be? I was thinking of trying to hack together a...

Make sure that you're using Theano 0.9.0 (`python -c 'import theano; print(theano.__version__)'` should print `0.9.0.something`). Make sure you've followed all of the [Theano installation instructions](http://deeplearning.net/software/theano/install.html). Specifically, if you're using a...

Hmm. It should be able to do the isolation process in ~30s or so, at least on a GPU. Are you using python 3? I've only tested on python 3....

You can configure the file parsing code however you want for your dataset by messing with the functions at the top of `data.py`. Some examples of valid filenames for the...

It's an out-of-memory error; the code currently tries to process songs all at once (rather than splitting them up and processing segments individually), and that's a problem for long songs!...

The network is fully convolutional, so there's not much of a difference between running it on segments and running it on the whole thing (the one possible difference is artifacts...

Ah, that might be numpy being more strict than the version I was using. Can you try adding `int()` around the two `float()` calls? Like: ``` newY = int(ceil(spectrogram.shape[1] /...

The neural network architecture was roughly based off of pix2pix https://arxiv.org/pdf/1611.07004.pdf, although it's fairly generic. There are probably better architectural choices–I didn't do a thorough hyperparameter sweep! The input shape...

The model in this project uses a regular stft (and, as mentioned earlier, only the amplitude).

I talk a bit about my data collection process [here](https://www.reddit.com/r/MachineLearning/comments/66j2i4/p_isolating_vocals_from_music_with_a_convnet/dgkm2nc/). I think you can probably get away with much less; the only really important part about the data is that...