Jurafsky & Martin is the canonical textbook but it's pre-deep learning. Check out the wavenet paper[0], and Alan Black's Festvox[1]. [0] https://deepmind.com/b

Jurafsky & Martin is the canonical textbook but it's pre-deep learning. Check out the wavenet paper[0], and Alan Black's Festvox[1]. [0] https://deepmind.com/blog/wavenet-generative-model-raw-audio/ https://deepmind.com/blog/wavenet-generative-model-raw-audio... [1] http://www.festvox.org/ http://www.festvox.org/