Benchmarking Generative Latent Variable Models for Speech
This work addresses the gap in speech generation for researchers, though it is incremental as it adapts an existing method to a new domain.
The paper tackled the problem of stochastic latent variable models underperforming compared to deterministic models in speech generation by benchmarking temporal LVMs and adapting the Clockwork VAE to speech, finding it outperformed previous LVMs and reduced the performance gap.
Stochastic latent variable models (LVMs) achieve state-of-the-art performance on natural image generation but are still inferior to deterministic models on speech. In this paper, we develop a speech benchmark of popular temporal LVMs and compare them against state-of-the-art deterministic models. We report the likelihood, which is a much used metric in the image domain, but rarely, or incomparably, reported for speech models. To assess the quality of the learned representations, we also compare their usefulness for phoneme recognition. Finally, we adapt the Clockwork VAE, a state-of-the-art temporal LVM for video generation, to the speech domain. Despite being autoregressive only in latent space, we find that the Clockwork VAE can outperform previous LVMs and reduce the gap to deterministic models by using a hierarchy of latent variables.