"As deep learning continues to prove successful in audio with tasks like speech recognition, enhancement, and synthesis, its application in audio effects is no exception. Neural audio effects involve the use of neural network components to construct systems that manipulate audio signals, often with the goal of emulating analog audio effects. However, neural audio effects also open the door to constructing novel effects that provide new methods for shaping sound that extend beyond traditional signal processing. In this talk, we will outline recent advancements in neural audio effects, their application in emulating analog effects, and draw connections between neural network architectures and signal processing. Then, we will introduce our recent work in extending neural audio effects to design novel effects and discuss emerging topics and open problems in this area of research."
Christian's website: https://www.christiansteinmetz.com