In this paper we study neural network models to capture intonation patterns of speech in Indian languages. We examine the performance of neural networks and support vector machines (SVM) for this purpose. Modeling the intonation pattern is the task of predicting the sequence of fundamental frequency ( ) values for the sequence of syllables in the given text. Analysis is performed on broadcast n...

the source of energy for playing aerophones is the air that is conducted through the instrument via the lips of the player. despite the fact that the function of the respiratory system is consistent among all players, the method of blowing varies from one instrument to another. every instrument, and more specifically, every player must find the most appropriate method of utilizing the lips vis-...

We examined pitch patterns within and across intonational phrases of Japanese read aloud by native and non-native (Mandarin Chinese) speakers. Japanese speakers change pitch ranges for each intonational phrase. The relative pitch ranges of neighboring intonational phrases indicate which intonational phrase belongs to which intonational phrase group. Chinese speakers are unable to acoustically c...

This paper describes a computer program for synthesizing fundamental frequency (F0} contours for English, and sketches the theory which underlies it. The F0 contour is described as a series of targets within an envelope specifying F0 range; the F0 contour between targets is computed by transition rules. The use of nonmonotonic transitions permits a sparser specification of the contour than has ...

In this paper we introduce the first steps towards a new datadriven method for extraction of intonation events that does not require any prerequisite prosodic labelling. Provided with data segmented on the syllable constituent level it derives local and global contour classes by stylisation and subsequent clustering of the stylisation parameter vectors. Local contour classes correspond to pitch...

Fo generation is an effective instrument for identifying intonational foreign accent (IFA), as it can be applied to determine the relevance of observed deviations in the intonation patterns produced by non-native speakers of a language, This study aims to show that the method of intonation description that is used and that also serves as the basis for F0 generation determines the form of the ob...

The paper describes an interface between generator and synthesizer of the German language concept-to-speech system VieCtoS. It discusses phenomena in German intonation that depend on the interaction between grammatical dependencies (projection of information structure into syntax) and prosodic context (performance-related modiications to intonation patterns). Phonological processing in our syst...

PURPOSE Previous work on the temporal coordination between gesture and speech found that the prominence in gesture coordinates with speech prominence. In this study, the authors investigated the anchoring regions in speech and pointing gesture that align with each other. The authors hypothesized that (a) in contrastive focus conditions, the gesture apex is anchored in the intonation peak and (b...

The standard evaluation of intonation models is by means of non-referenced subjective tests (pair or MOS) in which subjects rate the quality or compare different samples without any explicit reference. These tests are usually conducted on an isolated sentence basis. However, for a single sentence, with no contextual information, there are multiple valid intonations. A subject’s preference over ...