Autoregressive Models for Statistical Parametric Speech Synthesis

By Matt Shannon, Heiga Zen, William J. Byrne. 2013

68 citations recorded by the source corpus.

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

1. Dong Yu
2. Li Deng
3. Jinyu Li
4. Yoshua Bengio
5. Andrew W. Senior
6. Heiga Zen
7. Hasim Sak
8. Tara N. Sainath

In Signal Processing

1. Dong Yu
2. Li Deng
3. Jinyu Li
4. Heiga Zen
5. Hasim Sak
6. Andrew W. Senior
7. Tara N. Sainath
8. Yifan Gong

Citations in the corpus

• Deep Learning: Methods and Applications (2014, background): “Similar trajectory HMMs also form the basis for parametric speech synthesis [228, 326, 439, 438].”
• A tutorial survey of architectures, algorithms, and applications for deep learning (2014, background): “Similar trajectory HMMs also form the basis for parametric speech synthesis [62–66].”
• Deep mixture density networks for acoustic modeling in statistical parametric speech synthesis (2014, background): “Statistical parametric speech synthesis (SPSS) [1] based on hidden Markov models (HMMs) [2] offers various advantages over concatenative speech synthesis [3].”
• Statistical Parametric Speech Synthesis: A review (2015, methodology): “The LGLAR HMM, like the standard HMM synthesis framework, has certain parameters such as MDL tuning factor LGLAR HMM is capable of producing speech that is as natural as that of the standard HMM synthesis framework with its conventional settings, but not as natural as the trajectory HMM[6].”
• Connectionist multivariate density-estimation and its application to speech synthesis (2016, background): “Autoregressive HMMs are a promising research direction for speech synthesis (Shannon et al., 2013), but in preliminary experiments training fully-autoregressive RNADEs (autoregressive”
• Parameter Generation Algorithms for Text-To-Speech Synthesis with Recurrent Neural Networks (2018, background): “Yet another approach, that affects continuity of generated trajectories, is including autoregressive output layers into acoutic model of text-to-speech system [10, 11, 15].”
• Unidirectional long short-term memory recurrent neural network with recurrent output layer for low-latency speech synthesis (2015, methodology): “Use an acoustic model allowing streaming inference such as autoregressive hidden Markov models (AR-HMMs) [24].”