Rule-Based Storytelling Text-to-Speech (TTS) Synthesis

Ramli, Izzad; Seman, Noraini; Ardi, Norizah; Jamil, Nursuriati

doi:10.1051/matecconf/20167704003

Cited by 5 publications

(1 citation statement)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, the reading rate estimates we obtained from TTS are independent of discourse-related effects. In fact, because modern TTS systems only account for local effects, TTS researchers have tried to modify TTS systems for storytelling contexts to generate digital speech that accounts for emotions, characters, and other discourse-related features (Delmonte & Tripodi, 2015; Doukhan et al, 2011; Ramli et al, 2016; Theune et al, 2006). Nevertheless, investigating discourse-related features as a source of text-based ORF variability in fiction books like HP1 is a fruitful direction for future work.…”

Section: Discussionmentioning

confidence: 99%

Beyond text complexity: Production-related sources of text-based variability in oral reading fluency.

Liceralde¹,

Loukina²,

Klebanov³

et al. 2022

Journal of Educational Psychology

View full text Add to dashboard Cite

Variability in oral reading fluency (ORF), an indicator of foundational reading skills, has been linked to characteristics of texts. Such text-based variability in ORF has been traditionally attributed to text complexity, but substantial text-based variability has still been observed after accounting for text complexity. We consider that oral reading requires pronouncing the text aloud, which makes it subject to the same articulatory and prosodic constraints as other types of speech productions. Thus, texts with similar levels of complexity may still differ in expected durations when read aloud because of the texts’ segmental and prosodic structure, leading to differences in reading rate. We propose that these production-related effects are also important sources of text-based ORF variability. Data from upper elementary school students in the United States reading a large variety of passages from a popular fiction book showed that a composite measure of production-related effects (i.e., reading rate estimates derived from a text-to-speech synthesis system) explained a substantial amount of text-based ORF variability. Follow-up exploratory analyses indicated that these production-related effects are robust. Because text complexity metrics consist of features that also tap into production constraints, our results motivate an updated interpretation of text complexity effects on ORF and highlight the importance of accounting for production-related effects on ORF, which remain to be acknowledged in the ORF literature as potential sources of text-based variability.

show abstract

Section: Discussionmentioning

confidence: 99%

Beyond text complexity: Production-related sources of text-based variability in oral reading fluency.

Liceralde¹,

Loukina²,

Klebanov³

et al. 2022

Journal of Educational Psychology

View full text Add to dashboard Cite

show abstract

The Information Structure–prosody interface in text-to-speech technologies. An empirical perspective

Domínguez

Farrús

Wanner

2021

Corpus Linguistics and Linguistic Theory

View full text Add to dashboard Cite

The correspondence between the communicative intention of a speaker in terms of Information Structure and the way this speaker reflects communicative aspects by means of prosody have been a fruitful field of study in Linguistics. However, text-to-speech applications still lack the variability and richness found in human speech in terms of how humans display their communication skills. Some attempts were made in the past to model one aspect of Information Structure, namely thematicity for its application to intonation generation in text-to-speech technologies. Yet, these applications suffer from two limitations: (i) they draw upon a small number of made-up simple question-answer pairs rather than on real (spoken or written) corpus material; and (ii) they do not explore whether any other interpretation would better suit a wider range of textual genres beyond dialogs. In this paper, two different interpretations of thematicity in the field of speech technologies are examined: the state-of-art binary (and flat) theme-rheme, and the hierarchical thematicity defined by Igor Mel’čuk within the Meaning-Text Theory. The outcome of the experiments on a corpus of native speakers of US English suggests that the latter interpretation of thematicity has a versatile implementation potential for text-to-speech applications of the Information Structure–prosody interface.

show abstract

Speech Synthesis for Conversational News Contents Delivery

Takatsu

Fukuoka

Fujie

et al. 2019

Transactions of the Japanese Society for Artificial Intelligenc

View full text Add to dashboard Cite

keywords: conversational speech synthesis, DNN-based speech synthesis, paragraph-based speech synthesis, pause length estimation, prominence SummaryWe have been developing a speech-based "news-delivery system", which can transmit news contents via spoken dialogues. In such a system, a speech synthesis sub system that can flexibly adjust the prosodic features in utterances is highly vital: the system should be able to highlight spoken phrases containing noteworthy information in an article; it should also provide properly controlled pauses between utterances to facilitate user's interactive reactions including questions. To achieve these goals, we have decided to incorporate the position of the utterance in the paragraph and the role of the utterance in the discourse structure into the bundle of features for speech synthesis. These features were found to be crucially important in fulfilling the above-mentioned requirements for the spoken utterances by the thorough investigation into the news-telling speech data uttered by a voice actress. Specifically, these features dictate the importance of information carried by spoken phrases, and hence should be effectively utilized in synthesizing prosodically adequate utterances. Based on these investigations, we devised a deep neural network-based speech synthesis model that takes as input the role and position features. In addition, we designed a neural network model that can estimate an adequate pause length between utterances. Experimental results showed that by adding these features to the input, it becomes more proper speech for information delivery. Furthermore, we confirmed that by inserting pauses properly, it becomes easier for users to ask questions during system utterances. * 1 http://www.apple.com/jp/ios/siri/ * 2 https://www.softbank.jp/robot/consumer/ products/ 15, Yoshino 15][ 18a]

show abstract

Rule-Based Storytelling Text-to-Speech (TTS) Synthesis

Cited by 5 publications

References 11 publications

Beyond text complexity: Production-related sources of text-based variability in oral reading fluency.

Beyond text complexity: Production-related sources of text-based variability in oral reading fluency.

The Information Structure–prosody interface in text-to-speech technologies. An empirical perspective

Speech Synthesis for Conversational News Contents Delivery

Contact Info

Product

Resources

About