A Template-Based Approach for Speech Synthesis Intonation Generation Using LSTMs

By Srikanth Ronanki, Gustav Eje Henter, Zhizheng Wu, Simon King. 2016

19 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Model Architectures, Speech & Audio, 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. Sercan Ö. Arik
2. G. Diamos
3. Xu Tan
4. Tao Qin
5. Tie-Yan Liu
6. Adam Coates
7. John Miller
8. Axel Röbel

In Model Architectures

1. Adam Coates
2. John Miller
3. Axel Röbel
4. Yongguo Kang
5. Mohammad Shoeybi
6. G. Diamos
7. Carl Robinson
8. Nicolas Obin

In Speech & Audio

1. Sercan Ö. Arik
2. G. Diamos
3. Adam Coates
4. John Miller
5. Axel Röbel
6. Yongguo Kang
7. Carl Robinson
8. Nicolas Obin

In Signal Processing

1. Adam Coates
2. John Miller
3. Axel Röbel
4. Yongguo Kang
5. G. Diamos
6. Carl Robinson
7. Nicolas Obin
8. Shubho Sengupta

Citations in the corpus

• Deep Voice: Real-time Neural Text-to-Speech (2017, background): “, 2015; Yao & Zweig, 2015), phoneme duration prediction models (Zen & Sak, 2015), fundamental frequency prediction models (Pascual & Bonafonte, 2016; Ronanki et al., 2016), and audio synthesis models (van den Oord et al.”
• Median-based generation of synthetic speech durations using a non-parametric approach (2016, unknown): “It is suspected that generating durations and fundamental frequency contours that are jointly appropriate may be of importance for synthetic speech prosody; cf. [22, 23].”
• Sequence-to-sequence Modelling of F0 for Speech Emotion Conversion (2019, methodology): “A standard RNN can be extended to make both the hidden state and the output values recurrent [15], and separate LSTMs can be used for predicting phone durations and the other acoustic features [19].”