Evaluating generative audio systems and their metrics

Ashvala Vinay, Alexander Lerch

Recent years have seen considerable advances in audio synthesis with deep generative models. However, the state-of-the-art is very difficult to quantify; different studies often us…