On the training aspects of Deep Neural Network (DNN) for parametric TTS synthesis

By Yao Qian, Yuchen Fan, Wenping Hu, Frank K. Soong. 2014

220 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. Junichi Yamagishi
2. Xu Tan
3. Tao Qin
4. Tie-Yan Liu
5. Frank K. Soong
6. Lei Xie
7. Heiga Zen
8. Dong Yu

In Signal Processing

1. Junichi Yamagishi
2. Xu Tan
3. Frank K. Soong
4. Tao Qin
5. Tie-Yan Liu
6. Heiga Zen
7. Lei Xie
8. Simon King

Citations in the corpus

• A Survey on Neural Speech Synthesis (2021, methodology): “From 2010s, neural network-based speech synthesis [419, 278, 76, 417, 369, 187, 248, 376] has gradually become the dominant methods and achieved much better voice quality.”
• Fine-Grained Emotion Strength Transfer, Control and Prediction for Emotional Speech Synthesis (2020, background): “Thanks to the rapid development of deep learning, speech synthesis has been significantly advanced [1, 2, 3].”
• Fast, Compact, and High Quality LSTM-RNN Based Statistical Parametric Speech Synthesizers for Mobile Devices (2016, background): “ANN-based acoustic models offer an efficient and distributed representation of complex dependencies between linguistic and acoustic features [21, 22] and have shown the potential to produce natural sounding synthesized speech [2, 4, 7–9].”
• Composition of Deep and Spiking Neural Networks for Very Low Bit Rate Speech Coding (2016, background): “All these conditions are exposed even more in speech coders operating at VLBR speech coding of the order of hundreds of bit/s.”
• From Eliza to XiaoIce: challenges and opportunities with social chatbots (2018, background): “, 2013, 2015), speech recognition and synthesis (Hinton et al., 2012; Deng et al., 2013; Xiong et al., 2016; Qian et al., 2014; van den Oord et al., 2016), computer vision (Krizhevsky et al.”
• Multi-output RNN-LSTM for multiple speaker speech synthesis with α-interpolation model (2016, background): “In the case of speech synthesis, many works included DNNs and DBNs to perform acoustic mappings and prosody prediction [2, 3, 4], Also, Recurrent Neural Networks (RNNs) and their variants, like the Long Short Term Memory (LSTM) architecture [5], have leveraged completely the sequences processing and”
• Towards Multi-Scale Style Control for Expressive Speech Synthesis (2021, background): “In addition, in order to regulate the temporal granularity of the convolution output closer to human vocal perception, the filter strides of 6 convolution layers are set as [2, 1, 2, 1, 2, 2].”
• Spoofing and countermeasures for speaker verification: A survey (2015, background): “In addition to the four major approaches, inspired by advances in deep neural network (DNN)-based speech recognition (Hinton et al., 2012), new data-driven, DNN-based approaches have also been actively investigated (Zen et al., 2013; Ling et al., 2013; Lu et al., 2013; Qian et al., 2014).”
• Unidirectional long short-term memory recurrent neural network with recurrent output layer for low-latency speech synthesis (2015, background): “Introduction of the deep neural network (DNN) [3], which is a feed-forward artificial neural network with many hidden layers, has opened a new research direction for acoustic modeling in SPSS [4] [5] [6] [7].”
• Speaker and language factorization in DNN-based TTS synthesis (2016, methodology): “Deep Neural Networks (DNNs) have advanced parametric Text-to-Speech (TTS) synthesis to a new frontier [2, 3, 4, 5, 6, 7, 8, 9, 10].”
• Model architectures to extrapolate emotional expressions in DNN-based text-to-speech (2021, background): “Recently, deep neural network (DNN)-based TTS has been intensively investigated, and the results demonstrated that DNN-based TTS can outperform hidden Markov model (HMM)-based TTS in the quality and naturalness of synthesized speech [1, 2, 3, 4].”
• Multi-speaker modeling and speaker adaptation for DNN-based TTS synthesis (2015, background): “Deep neural networks (DNNs) has shown its great power for acoustic modeling in TTS synthesis.”
• Singing Voice Synthesis Based on Generative Adversarial Networks (2019, background): “In recent years, deep neural networks (DNNs) have attained significant improvement in various machine learning areas such as image recognition [4, 5], speech recognition [6], and speech synthesis [7, 8].”
• Review of end-to-end speech synthesis technology based on deep learning (2021, methodology): “TTS methods based on deep learning have been widely researched [52, 131, 168, 247].”
• A DNN-based Mandarin-Tibetan cross-lingual speech synthesis (2018, methodology): “The interaction between input context and output acoustic features is modeled by a DNN In [10], which can address some limitations of conventional HMM-based speech synthesis approaches.”
• Deep Feed-Forward Sequential Memory Networks for Speech Synthesis (2018, methodology): “rms— FSMN, Deep-FSMN, parametric speech synthesis, Text-to-Speech 1. INTRODUCTION In recent years, Deep Neural Networks (DNNs) have became dominant in the parametric speech synthesis back-end modeling[1, 2]. Compared with conventional HMM-GMM methods training at state level, neural network based met”
• Phonological vocoding using artificial neural networks (2015, background): “Index Terms— Parametric vocoding, low bit rate speech coding, phonology.”
• A probabilistic interpretation for artificial neural network-based voice conversion (2015, methodology): “A notable example is maximum likelihood parameter generation (MLPG) algorithm [2, 13] for ANN-based VC [7, 8] and deep neural network (DNN)-based speech synthesis (SS) [15, 16], with the aim of overcoming the discontinuity problem in the conventional ANN-based parameter generation process.”