WaveNet

Deep Learning for Computer Vision (Andrej Karpathy, OpenAI)

Lex Fridman

A deep generative model from DeepMind that uses dilated convolutions for processing audio, mentioned as an example for handling large contexts.