Neural Instrument Cloning from very few samples

[erl-j]

Hello!

I want to share some results I’ve obtained from applying some model adaptation strategies from speech synthesis to ddsp. I have audio examples and a colab to play around with here.

Have a nice day!