For September 10
-If you want to work on the deep learning components read this - http://karpathy.github.io/2015/05/21/rnn-effectiveness/
-at the moment this is how both melody and chord generation
-Shannon - begin on chord implementation, present what you did last semester
-Johnathan - be ready to present on what you worked on last semester
-Adam - Check out MeloSpy - https://jazzomat.hfm-weimar.de/download/download.html and Music21 http://web.mit.edu/music21/ You may already know them, if not try and get both running.
-Rob - background prep in deep learning. I'm pretty happy with using an LSTM for chord generation, but open to new models for melody (also chords). Check out - https://skymind.ai/wiki/generative-adversarial-network-gan
--- If anyone wants to try deep learning work on a good GPU, we have two linux machines that you could use here, one with dual Titan-Vs and one with dual 1080s. Just let me know and I'll get you access.
---
Chord Generation
RNN For Generation
http://karpathy.github.io/2015/05/21/rnn-effectiveness/
Java Version of Chord Analysis
https://github.com/suinyun/vip1_robotic_musicianship
Rules
Tonal/Variation/Stepping Through
14th of September:
Goal:
Method:
Results:
Rob:
Read Papers, Read textbook chapter (Richard will put up list)
GAN's
Teaching everyone GAN's next week
Shannon:
Chord Classifier to Python
Johnathan:
Adam
Resources