notes.txt

Getting Data:

Download as a mp3, wav file (Audio only)
Python uses wav files
	youtube-dl -f bestaudio --extract-audio --audio-format mp3 --audio-quality 0 <Video-URL>
	youtube-dl -f bestaudio --extract-audio --audio-format wav --audio-quality 0 <Video-URL>

TODO:
	analyze left and right channels of audio 2D (taki_taki)
	

Pretrained models:
	https://github.com/hindupuravinash/the-gan-zoo

Evaluate output:
	Play in terminal aplay <AUDIO-FILE>

Explaining keras inputs:
	https://stackoverflow.com/questions/44747343/keras-input-explanation-input-shape-units-batch-size-dim-etc	

Start with:
For classification of music into labels
	https://www.analyticsvidhya.com/blog/2017/08/audio-voice-processing-deep-learning/

******** Try taking a look at DLND on how to generate new episodes

Open Source:
	https://github.com/chrisdonahue/wavegan
	https://magenta.tensorflow.org/

Siraj Raval:
AI for Music Composition
https://www.youtube.com/watch?v=NS2eqVsnJKo

Currently working off this:
	https://github.com/usernaamee/audio-GAN/blob/master/simple-audio-GAN.py --x
	https://blog.keras.io/building-autoencoders-in-keras.html
try building to:
	https://github.com/HackerPoet/Composer

Wav files are 2-D (left channel, right channel)