Comment by prashp

Is it possible to model reverb using a neural network (e.g. wavenet or LSTMs) for real-time use? Is this what something like Neural DSP is doing under the hood?