What books describe convolutional neural networks for DSP?

Question

(note: I read the site policy and meta questions about whether such a question is ok. I will be specific)

I am looking for books that specifically cover using convolutional neural networks (CNN) for audio processing for signal detection, including the preprocessing steps for the audio and the network architecture. I need direction on preprocessing like the best sample rates, time frames for STFT/spectrograms and image size for a CNN. Do such books exist? I couldn't find any on Amazon.

I would also appreciate links to the books on Amazon (or elsewhere) and links to github, if possible.

Thank you!

Answer 1

To my knowledge there might not be good books on it already.

I'll assume you have enough background in classical DSP and CNNs ( and NNs in general): go read the papers that things like WaveRNN, torchaudio etc cite.

Papers, not books, because

1. the field might still be too young, so knowledge isn't mature enough to warrant writing a book on it instead of doing forward research in the same time, and
2. your restriction to care about convolutional neural networks only might be too impractical to fill a book, but certainly has been investigated in papers.