GAN Vocoder: Multi-Resolution Discriminator Is All You Need
arXiv:2103.05236
Abstract
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.
arXiv commentAccepted to Interspeech 2021
References in corpus (8)
- Decoupled Weight Decay Regularization
- MelGAN: Generative Adversarial Networks for Conditional Waveform Synthesis
- Adversarial Audio Synthesis
- Very Deep VAEs Generalize Autoregressive Models and Can Outperform Them on Images
- Effective Use of Variational Embedding Capacity in Expressive End-to-End Speech Synthesis
- Universal MelGAN: A Robust Neural Vocoder for High-Fidelity Waveform Generation in Multiple Domains
- Improved parallel WaveGAN vocoder with perceptually weighted spectrogram loss
- Axial Residual Networks for CycleGAN-based Voice Conversion