Pipeline for VA Modelling with Physics-Informed Machine Learning

Explore a cutting-edge approach to virtual analog modeling in this conference talk from the Audio Developer Conference. Delve into a pipeline that combines Wave Digital Filters with neural networks for circuit modeling. Learn how the Differentiable Wave Digital Filters library was utilized to train a real-time deployable model of a diode clipper circuit. Discover the advantages of this methodology, including higher accuracy and comparable computation time to traditional white-box models. Gain insights from Christopher Johann Clarke, a PhD candidate specializing in multiphysics and AI/machine learning, as he presents this innovative approach to audio circuit modeling.