A Deep Generative Architecture for Postfiltering in Statistical Parametric Speech Synthesis

By Linghui Chen, Tuomo Raitio, Cassia Valentini-Botinhao, Zhenhua Ling, Junichi Yamagishi. 2015

50 citations recorded by the source corpus.

The generated speech of hidden Markov model (HMM)-based statistical parametric speech synthesis still sounds “muffled.” One cause of this degradation in speech quality may be the loss of fine spectral structures. In this paper, we propose to use a deep generative architecture, a deep neural network (DNN) generatively trained, as a postfilter. The network models the conditional probability of the spectrum of natural speech given that of synthetic speech to compensate for such gap between synthetic and natural speech. The proposed probabilistic postfilter is generatively trained by cascading two restricted Boltzmann machines (RBMs) or deep belief networks (DBNs) with one bidirectional associative memory (BAM). We devised two types of DNN postfilters: one operating in the mel-cepstral domain and the other in the higher dimensional spectral domain. We compare these two new data-driven postfilters with other types of postfilters that are currently used in speech synthesis: a fixed mel-cepstral based postfilter, the global variance based parameter generation, and the modulation spectrum-based enhancement. Subjective evaluations using the synthetic voices of a male and female speaker confirmed that the proposed DNN-based postfilter in the spectral domain significantly improved the segmental quality of synthetic speech compared to that with conventional methods.

Subfields: Artificial Intelligence, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Signal Processing

Citations in the corpus

Generative adversarial network-based postfilter for statistical parametric speech synthesis (2016, methodology): “A previous work [14, 15, 16] tried to represent it using a learning-based approach similar to ours, but they assumed its distribution as Gaussian (Chen et al.”
Speech Enhancement for a Noise-Robust Text-to-Speech Synthesis System Using Deep Recurrent Neural Networks (2016, background): “The structure could also be seen as a pre-filter (as opposed to a postfilter that acts at the vocoder level at generation time [22]) and in that sense could potentially be used to minimise synthesis errors as well as enhancement errors.”