Text-to-Speech Synthesis

By Paul Taylor. 2009

662 citations recorded by the source corpus.

How researchers see this paper

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

In Speech & Audio

Citations in the corpus

NaturalSpeech 2: Latent Diffusion Models are Natural and Zero-Shot Speech and Singing Synthesizers (2023, background): “Text-to-speech (TTS) [1, 2] aims to synthesize natural and human-like speech with both good quality and diversity.”
A Survey on Neural Speech Synthesis (2021, background): “Text to speech (TTS), also known as speech synthesis, which aims to synthesize intelligible and natural speech from text [340], has broad applications in human communication [1] and has long been a research topic in artificial intelligence, natural language and speech processing [290, 222, 144].”
Sample Efficient Adaptive Text-to-Speech (2018, neutral): “We show this for adaptive TTS (Dutoit, 1997; Taylor, 2009).”
Deep Voice 3: 2000-Speaker Neural Text-to-Speech (2017, background): “1 INTRODUCTION Artificial speech synthesis, also called text-to-speech (TTS), is traditionally done with complex multi-stage hand-engineered pipelines (Taylor, 2009).”
End-to-End Adversarial Text-to-Speech (2020, neutral): “Typical TTS pipelines consist of a number of stages trained or designed independently – e.g. text normalisation, aligned linguistic featurisation, mel-spectrogram synthesis, and raw audio waveform synthesis [56].”
Neural Source-Filter Waveform Models for Statistical Parametric Speech Synthesis (2019, background): “T EXT-TO-SPEECH synthesis (TTS), a technology that converts a text string into a speech waveform [1], has rapidly advanced with the help of deep learning.”
Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion Models (2023, background): “…crop it to a mel-spectrogram of size 80 × 624 ; 3) non-standard words (e.g., abbreviations, numbers, and currency expressions) and semiotic classes (Taylor, 2009) (text tokens that represent particular entities that are semantically constrained, such as measure phrases, addresses, and dates) are…”
Improving Prosody with Linguistic and Bert Derived Features in Multi-Speaker Based Mandarin Chinese Neural TTS (2020, background): “A classical SPSS (Statistical Parametric Speech Synthesis) system [1] usually consists of several different components, including: a frontend-module (convert text to linguistic features), an acoustic model (map linguistic features to acoustic features) and a vocoder (generate speech waveforms from a”
Speech Synthesis Based on Hidden Markov Models (2013, methodology): “, [1] and [24]–[35], have also been proposed.”
Intent transfer in speech-to-speech machine translation (2012, background): “These aspects broadly fall under the 'augmentative' and 'affective' parts of prosody, the extra information in speech, to ensure that the intended message is unambiguously decoded by the listeners [1].”
Natural TTS Synthesis by Conditioning Wavenet on MEL Spectrogram Predictions (2017, background): “Generating natural speech from text (text-to-speech synthesis, TTS) remains a challenging task despite decades of investigation [1].”
DeviceTTS: A Small-Footprint, Fast, Stable Network for On-Device Text-to-Speech (2020, neutral): “These two techniques are based on complex multi-stage hand-engineered pipelines [1], and the synthetic speech often sounds mufﬂed and unnatural compared to the human recording.”
Postfilters to Modify the Modulation Spectrum for Statistical Parametric Speech Synthesis (2016, methodology): “The synthetic speech parameter sequence is generated by the following algorithms.”
Parallel Neural Text-to-Speech (2019, background): “Traditional TTS systems are based on multi-stage hand-engineered pipelines (Taylor, 2009).”
The STC text-to-speech system for Blizzard Challenge 2019 (2019, background): “The task may also require time-consuming labeling of speech corpora by hand [1].”
A Neural Parametric Singing Synthesizer (2017, background): “Additionally, phonetics can also cause inflections in F0, so-called microprosody [16].”
Forward–Backward Decoding Sequence for Regularizing End-to-End TTS (2019, background): “[17]–[20] are proposed to simplify traditional TTS pipeline [3], [6], [11], [12] with a single neural network.”