Audio Processing in Keras with Kapre

Traditionally in audio processing, features (e.g., mel spectrograms) are pre-computed and then fed to machine learning models. A different approach to audio processing is to embed audio transformations as part of a deep learning model and leverage GPU for calculations. This method has some clear advantages: DSP parameters used in the transformations can be optimised at training time. Deployment of the model is easier. Less dependencies to handle. In my last video, you can learn more about this approach and especially about Kapre. Kapre is an audio signal processing library for Keras developed by Keunwoo Choi. Kapre’s documentation: Kapre’s GitHub: ================= Interested in hiring me as a consultant/freelancer? Join The Sound Of AI Slack community: Connect with Vale