FloWaveNet : A Generative Flow for Raw Audio
This addresses the high inference time bottleneck in practical text-to-speech applications, offering a more efficient alternative to existing parallel models.
The paper tackles the problem of slow inference in WaveNet-based text-to-speech by proposing FloWaveNet, a flow-based generative model that achieves real-time audio synthesis with clarity comparable to previous two-stage models, using only single-stage training and a single maximum likelihood loss.
Most modern text-to-speech architectures use a WaveNet vocoder for synthesizing high-fidelity waveform audio, but there have been limitations, such as high inference time, in its practical application due to its ancestral sampling scheme. The recently suggested Parallel WaveNet and ClariNet have achieved real-time audio synthesis capability by incorporating inverse autoregressive flow for parallel sampling. However, these approaches require a two-stage training pipeline with a well-trained teacher network and can only produce natural sound by using probability distillation along with auxiliary loss terms. We propose FloWaveNet, a flow-based generative model for raw audio synthesis. FloWaveNet requires only a single-stage training procedure and a single maximum likelihood loss, without any additional auxiliary terms, and it is inherently parallel due to the characteristics of generative flow. The model can efficiently sample raw audio in real-time, with clarity comparable to previous two-stage parallel models. The code and samples for all models, including our FloWaveNet, are publicly available.