A Style Control Technique for HMM-Based Expressive Speech Synthesis

By Takashi Nose, Junichi Yamagishi, Takashi Masuko, Takao Kobayashi. 2007

155 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Experimental and Cognitive Psychology, 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 Experimental and Cognitive Psychology

In Signal Processing

Citations in the corpus

Towards End-to-End Prosody Transfer for Expressive Speech Synthesis with Tacotron (2018, background): “(Nose et al., 2007) proposes estimating the transformation matrix for a set of predefined style vectors.”
Statistical parametric speech synthesis using deep neural networks (2013, background): “This approach has various advantages over the concatenative speech synthesis approach [2], such as the flexibility to change its voice characteristics , [3] [4] [5] [6], small footprint [7] [8] [9], and robustness [10].”
Adapting and controlling DNN-based speech synthesis using input codes (2017, methodology): “, a different emotion) [1, 2]; 2) speaker interpolation, a technique to morph together several speakers or emotions by interpolating mean vectors and variance matrices of Gaussian distributions from representative acoustic models [3]; and 3) multiple regression HMMs (MR-HMMs) [4], a technique that m”
Statistical Parametric Speech Synthesis (2007, methodology): “…approach (Fujinaga et al., 2001) to HMM-based speech synthesis to control voice characteristics intuitively (Miyanaga et al., 2004; Nose et al., 2007b), where mean vectors of state-output distributions 9 were controlled with an - dimensional control vector, , as where is a…”
Sentence-level control vectors for deep neural network speech synthesis (2015, background): “In contrast to [3] and [17], however, we initialise our utterance representations randomly, and learn them in an entirely unsupervised fashion.”
Speech Synthesis Based on Hidden Markov Models (2013, methodology): “In [29] and [30], these auxiliary features for TTS are Fig.”
Hidden semi-Markov models (2010, unknown): “[132–134], Moore and Savic [125] and Wu et al.”
Generative adversarial network-based postfilter for statistical parametric speech synthesis (2016, background): “In the field of speech synthesis, statistical parametric speech synthesis based on hidden Markov models (HMMs) [1] or deep neural networks (DNNs) [2] has become popular since it provides various advantages over concatenative speech synthesis [3], such as the flexibility to control its voice characte”
Analysis of Speaker Adaptation Algorithms for HMM-Based Speech Synthesis and a Constrained SMAPLR Adaptation Algorithm (2009, methodology): “We have indeed already shown that the emotional expressions and speaking styles of synthetic speech can be easily reproduced, controlled, and transformed by using style modeling [13], model adaptation [14], model interpolation and model morphing [15], or multiple-regression HMMs [16].”
An investigation of multi-speaker training for wavenet vocoder (2017, background): “The main advantage of this approach is its high controllability, since it allows us to flexibly change the voice characteristics of synthesized speech by manipulating acoustic features [11, 12].”
Articulatory Control of HMM-Based Parametric Speech Synthesis Using Feature-Space-Switched Multiple Regression (2013, unknown): “In this method, a separate Gaussian mixture model (GMM) is introduced to model the articulatory space, and the regression matrices are estimated for each mixture component in this GMM instead of for each HMM state.”
Improving Unsupervised Style Transfer in end-to-end Speech Synthesis with end-to-end Speech Recognition (2018, background): “A speaker verification model [14, 15] can be pretrained to extract the speaker characteristics from the reference utterance [16].”
Postfilters to Modify the Modulation Spectrum for Statistical Parametric Speech Synthesis (2016, background): “This statistical modeling and generation framework make it possible to build small footprint synthesizers [21], adapt existing voices to other target voices using only a small amount of speech data [22], and ﬂexibly control voice characteristics of synthetic speech [23], [24].”
Uncovering Latent Style Factors for Expressive Speech Synthesis (2017, background): “[8] proposes estimating the transformation matrix for a set of predefined style vectors.”