Simultaneous Modeling of Spectrum, Pitch and Duration in HMM-Based Speech Synthesis

By Takayoshi Yoshimura, Keiichi Tokuda, Takashi Masuko, Takao Kobayashi, Tadashi Kitamura. 1999

How researchers see this paper

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

1. Xu Tan
2. Heiga Zen
3. Tao Qin
4. Li Deng
5. Junichi Yamagishi
6. Dong Yu
7. Keiichi Tokuda
8. Zhaoxi Mu

Citations in the corpus

• ELLA-V: Stable Neural Codec Language Modeling with Alignment-guided Sequence Reordering (2024, methodology): “Early methods were based on Statistical Parametric Speech Synthesis (SPSS) (Zen et al., 2009), typically involving complex components such as text analysis models, acoustic models, and vocoders ( e . g ., hidden Markov model(HMM) (Yoshimura et al., 1999) based).”
• Vocos: Closing the gap between time-domain and Fourier-based neural vocoders for high-quality audio synthesis (2023, methodology): “Within the scope of text-to-speech (TTS), concatenative synthesis (Moulines & Charpentier, 1990; Hunt & Black, 1996) and statistical parametric synthesis (Yoshimura et al., 1999) were the prevailing approaches.”
• A Survey on Neural Speech Synthesis (2021, methodology): “Statistic parametric synthesis [409, 350, 408, 418, 351] uses three basic modules, where text analysis module converts characters into linguistic features, and acoustic models generate acoustic features from linguistic features (where the target acoustic features are obtained through vocoder analysi”
• Listening while speaking: Speech chain by deep learning (2017, methodology): “In a similar direction, TTS technology development has gradually shifted from the foundation of a rule-based system using waveform coding and an analysis-synthesis method [7, 8] to a waveform unit concatenation approach [9, 10] and a more flexible approach using the statistical modeling of hidden sem”
• Modeling Spectral Envelopes Using Restricted Boltzmann Machines and Deep Belief Networks for Statistical Parametric Speech Synthesis (2013, methodology): “In the common spectral modeling methods, the probability density functions (PDF) of each HMM state is represented by a single Gaussian distribution with diagonal covariance matrix and the distribution parameters are estimated under the maximum likelihood (ML) criterion [2].”
• Modeling spectral envelopes using restricted Boltzmann machines for statistical parametric speech synthesis (2013, methodology): “The probability density functions (PDF) of each HMM state is commonly represented by a single Gaussian distribution [1].”
• Review of end-to-end speech synthesis technology based on deep learning (2021, methodology): “of TTS techniques from the formant-based methods [92, 97, 105, 106, 179, 224] to the unit selection-based waveform cascade methods [6, 24, 36, 59, 80, 143, 144], and to the hidden Markov model (HMM)-based statistical parametric speech synthesis (SPSS) methods [26, 94, 149, 178, 210, 241, 249, 250].”
• Product of Experts for Statistical Parametric Speech Synthesis (2012, methodology): “In a typical HMM-based statistical parametric speech synthesis system, the state durations are modeled explicitly by context-dependent single Gaussian distributions clustered by decision trees [38].”
• Effective Spectral and Excitation Modeling Techniques for LSTM-RNN-Based Speech Synthesis Systems (2017, background): “, a pulse or noise (PoN) [6], creating buzzy or whispery sounds.”
• Machine Speech Chain (2020, methodology): “…technology has gradually shifted from the foundation of a rule-based system using waveform coding and an analysis-synthesis method [7]– [9] to waveform unit concatenation [10], [11] and a more flexible approach using the statistical modeling of a hidden semi-Markov model-GMM (HSMM-GMM) [12], [13].”
• 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”
• A Template-Based Approach for Speech Synthesis Intonation Generation Using LSTMs (2016, unknown): “Conventional HMM+regression tree speech synthesis predicts frame-wise F0 from contextual linguistic features either using a multi-space probability distribution (MSD) [4] or by continuous F0 modelling [5].”
• Statistical Parametric Speech Synthesis (2007, methodology): “Although many research groups have contributed to progress in statistical parametric speech synthesis, the description given here is somewhat biased toward implementation on the HMM-based speech synthesis system (HTS) 1 (Yoshimura et al., 1999; Zen et al., 2007b) for the sake of logical coherence.”
• Articulatory Control of HMM-Based Parametric Speech Synthesis Using Feature-Space-Switched Multiple Regression (2013, unknown): “IN recent years, hidden Markov models (HMM) have been successfully applied to acoustic modelling for speech synthesis, and HMM-based parametric speech synthesis has become a mainstream speech synthesis method [1], [2].”
• Speech Synthesis Based on Hidden Markov Models (2013, background): “The number of downloads of HTS exceeds 10,000.”