Abstract | ||
---|---|---|
This paper introduces WaveGrad 2, a non-autoregressive generative model for text-to-speech synthesis. WaveGrad 2 is trained to estimate the gradient of the log conditional density of the waveform given a phoneme sequence. The model takes an input phoneme sequence, and through an iterative refinement process, generates an audio waveform. This contrasts to the original WaveGrad vocoder which conditions on mel-spectrogram features, generated by a separate model. The iterative refinement process starts from Gaussian noise, and through a series of refinement steps (e.g., 50 steps), progressively recovers the audio sequence. WaveGrad 2 offers a natural way to trade-off between inference speed and sample quality, through adjusting the number of refinement steps. Experiments show that the model can generate high fidelity audio, approaching the performance of a state-of-the-art neural TTS system. We also report various ablation studies over different model configurations. Audio samples are available at https://wavegrad.github.io/v2. |
Year | DOI | Venue |
---|---|---|
2021 | 10.21437/Interspeech.2021-1897 | Interspeech |
DocType | Citations | PageRank |
Conference | 1 | 0.36 |
References | Authors | |
0 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Nanxin Chen | 1 | 64 | 7.55 |
Yu Zhang | 2 | 160 | 63.25 |
Heiga Zen | 3 | 1922 | 103.73 |
Ron J. Weiss | 4 | 443 | 29.47 |
Mohammad Norouzi | 5 | 1212 | 56.60 |
Najim Dehak | 6 | 5 | 3.48 |
William Chan | 7 | 357 | 24.67 |