Do you work with audio data? If we add a Keras layer to convert raw audio to Mel spectrograms, would you be interested?

→ View original post on X — @fchollet