GAN Vocoder: Multi-Resolution Discriminator Is All You Need
This work provides a foundational insight for researchers and practitioners in speech synthesis, showing that incremental improvements in generator design may be unnecessary when using a robust multi-resolution discriminator.
The paper tackled the problem of identifying the key factor behind the success of GAN-based vocoders, finding that a multi-resolution discriminating framework, rather than specific architectural details, is sufficient for achieving state-of-the-art performance in text-to-speech synthesis, with all tested generators performing indistinguishably across perceptual metrics.
Several of the latest GAN-based vocoders show remarkable achievements, outperforming autoregressive and flow-based competitors in both qualitative and quantitative measures while synthesizing orders of magnitude faster. In this work, we hypothesize that the common factor underlying their success is the multi-resolution discriminating framework, not the minute details in architecture, loss function, or training strategy. We experimentally test the hypothesis by evaluating six different generators paired with one shared multi-resolution discriminating framework. For all evaluative measures with respect to text-to-speech syntheses and for all perceptual metrics, their performances are not distinguishable from one another, which supports our hypothesis.