Prosodic modeling with rich syntactic context in HMM-based Mandarin speech synthesis

Yu, Yansuo; Li, Dongchen; Wu, Xihong

doi:10.1109/chinasip.2013.6625313

Cited by 9 publications

(6 citation statements)

References 10 publications

(8 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A development set of 500 sentences is required to adjust the parameters. The possible values of low threshold are set to be (3,4,5,6), and the possible values of up thresholds are (7,8,9,10). On the development set, the best performance was obtained when low threshold and up threshold were 4 and 8 respectively.…”

Section: Resultsmentioning

confidence: 99%

“…They found that the level information of a word in the syntactic tree is one of the best five features for predicting prosodic phrase boundaries. The authors of [8] demonstrated that the syntactic structure information was able to significantly improve the duration generation in a speech synthesis system. Zhang et al noticed that the prosodic structure of a sentence is also a tree, and they proposed a method of converting a syntactic tree to a prosodic tree [5].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Mandarin Prosodic Phrase Prediction based on Syntactic Trees

Zhang¹,

Wu²,

Yang³

et al. 2016

9th ISCA Workshop on Speech Synthesis Workshop (SSW 9)

View full text Add to dashboard Cite

Prosodic phrases (PPs) are important for Mandarin Text-To-Speech systems. Most of the existing PP detection methods need large manually annotated corpora to learn the models. In this paper, we propose a rule based method to predict the PP boundaries employing the syntactic information of a sentence. The method is based on the observation that a prosodic phrase is a meaningful segment of a sentence with length restrictions. A syntactic structure allows to segment a sentence according to grammars. We add some length restrictions to the segmentations to predict the PP boundaries. An F-Score of 0.693 was obtained in the experiments, which is about 0.02 higher than the one got by a Conditional Random Field based method.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Mandarin Prosodic Phrase Prediction based on Syntactic Trees

Zhang¹,

Wu²,

Yang³

et al. 2016

9th ISCA Workshop on Speech Synthesis Workshop (SSW 9)

View full text Add to dashboard Cite

show abstract

“…For example, the silence between VP and NP is usually longer than the break between NP and NP. Researches also implied that longer duration of break often occurred at the boundary with higher level syntactic phrase [24].…”

Section: Syntactic Phrase and Dependence Featuresmentioning

confidence: 95%

Investigating Effect of Rich Syntactic Features on Mandarin Prosodic Boundaries Prediction

Che

Wen

et al. 2015

J Sign Process Syst

View full text Add to dashboard Cite

Previous researches indicated that syntactic information could improve the performance of automatic prosodic boundaries prediction. However, few researches focused on the usefulness of various syntactic features on different phrase boundaries prediction in detail, especially the verification on large-scale corpus. This paper investigates the effect of different syntactic feature combinations for phrase boundaries prediction based on a large-scale Mandarin corpus. Syntactic phrase structure and dependency relationship are both introduced and compared in the experiments. The evaluations of all the features in different time-span as well as feature importance are also carried out. The experimental results show that prediction models of prosodic word and prosodic phrase achieve the best performance with syntactic phrase and dependency features, while the models with dependency features outperform other models when predicting intonational phrase. Furthermore, the performance of intonational phrase is improved obviously by concerning syntactic global features. Meanwhile, the experiments also find that the relationship between syntactic local features and lower level prosodic boundaries: prosodic phrase and prosodic word, is closer.

show abstract

“…The phrase types have been proven related to break level many years ago [16]. For example, the length of break between VP and NP is always longer than the break between NP and NP.…”

Section: Syntactic Phrase Featuresmentioning

confidence: 99%

Investigating effect of rich syntactic features on Mandarin prosodic phrase boundaries prediction

Che¹,

Wen²,

Li³

et al. 2014

The 9th International Symposium on Chinese Spoken Language Processing

View full text Add to dashboard Cite

Previous researches indicated that the performance of automatic prosodic boundaries prediction benefited from syntactic information for Mandarin. However, few researches focused on the usefulness of various syntactic features on different phrase boundaries prediction detailly, especially on large-scale corpus. This paper investigates the effect of different syntactic features combinations for prosodic phrase boundaries prediction based on a large-scale Mandarin corpus. The experimental results shows that prediction models of prosodic word and prosodic phrase achieve best performance with syntactic phrase and dependency features. Different from other two types of boundaries, the models with dependency features outperform than models with other features when prediction object is intonational phrase.

show abstract

Prosodic modeling with rich syntactic context in HMM-based Mandarin speech synthesis

Cited by 9 publications

References 10 publications

Mandarin Prosodic Phrase Prediction based on Syntactic Trees

Mandarin Prosodic Phrase Prediction based on Syntactic Trees

Investigating Effect of Rich Syntactic Features on Mandarin Prosodic Boundaries Prediction

Investigating effect of rich syntactic features on Mandarin prosodic phrase boundaries prediction

Contact Info

Product

Resources

About