From information structure to intonation (original) (raw)

From Information Structure to Intonation: A Phonolo9ical Interface for Concept-to-Speech

The pal)er describes an interface between generator and synthesizer of tile 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 (performancerelated modifications to intonation patterns).

From Information Structure to Intonation: A Phonological Interface for Concept-to-Speech

2008

The pal)er describes an interface between gen-erator and synthesizer of tile German language concept-to-speech system VieCtoS. It discusses phenomena in German intonation that depend on the interaction between grammatical depen-dencies (projection of information structure into syntax) and prosodic context (performance-related modifications to intonation patterns). Phonological processing in our system con> prises segmental as well as suprasegmental di-mensions such as syllabification, modification of word stress positions, and a symbolic encoding of intonation. Phonological phenomena often touch upon more than one of these dimensions, so that mutual accessibility of the data struc-tures on each dimension had to be ensured. We present a linear representation of the multidimensional phonological data based on a straightforward linearization convention, which suffices to bring this conceptually multilinear data set under the scope of the well-known pro-cessing techniques for two-lev...

From Information Structure to Intonation: A Phonological

2002

Using two-level morphology as a generator-synthesizer interface for concept-to-speech

1995

In a project for the development of a concept-to-speech system for German, we apply extended two-level-morphology (Trost 1991) to provide a uni ed solution to the tasks of morphotactics, segmental (morpho)phonology, syllabi cation and assignment of stress. Starting from a lexeme-based lexicon, we show that a declarative two-levelimplementation of a single rule-corpus complemented with feature lters is su cient for a comprehensive account of the various mutual in uences holding between separate phonological dimensions in the phonology of German.

Using Two-Level Morphology as a Generator-Synthesizer Interface in Concept-to-Speech Generation

Integrating stress and intonation into a concept-to-speech system

Proceedings of the 13th conference on Computational linguistics -, 1990

The paper deals with the integration of intonation algorithms into a concept-to-speech system for German 1). The algorithm for computing the stress hierarchy of a sentence introduced by Kiparski (1973) and the theory of syntactic grouping for intonation patterns developed by Bierwisch (1973) have been studied extensively, but they have never been implemented in a concept-to-speech system like the one presented here. We describe the back end of this concept-to-speech system: The surface generator transfers a hierarchical dependency structure of a sentence into a phoneme string by traversing it in a recurs~ve-descent manner.

Matching a tone-based and tune-based approach to English intonation for concept-to-speech generation

Proceedings of the 18th conference on Computational linguistics -, 2000

Tlle paper describes the results of a comparison of two annotation systems for isstoslal;ion, the tone-based ToBI al)proach and the 1;unebased api)roach proposed by Systemic Functi(mal Grammar (SFO). The goal of this comparison is to detine a mapping between the two systems tbr the purpose of concept-to-speech generation of English. Since ToB: is widely used in Sl)eech synthesis and SFG is widely used in nal;ural language generation and oft~rs a linguistically motivated aecollnt of intonation, it; appears a promising step to comt)ine the two approaches for concept-to-speech. A corpus of English utterances has been analysed with both ~].~()13I and SFG categories; eomparison of the analysis results has lead to the identification of some basic equivalents between the two systems on which a mapping can be based.

Building Prosodic Structures in a Concept-to-Speech System

2003

The prosodic structure of utterances in terms of breaks and tones is a significant problem in speech synthesis. In this work we present the results from models used to predict accurate and realistic prosodic structures within the context of a Concept-to-Speech system for a virtual museum guide. We have used a Natural Language Generator system for providing error-free enriched linguistic information, such as syntax and Part-of-Speech, to a Speech Synthesizer. An XML annotation has been used as a means for this transfer of linguistic data. The annotated data was used to build classification trees for the prediction of prosodic phrase breaks, pitch accents and endtones (phrase accents and boundary tones). The annotation of utterances included segmental information, ToBI marks, syntax, grammar and some domain specific features such as new/given and phrase subject/object information. The linguistic nature of the domain allowed us to carefully select the set of features and the training conditions and also to utilize speech-oriented information from the written language produced by the Natural Language Generator component, such as evidence of stress and intonational focus. A speech corpus of 516 utterances has been used for training and evaluation purposes. To optimize the generated models, we used exhaustive training upon the domain data, achieving a correlation between the observed and the predicted elements of 97.286% for phrase breaks, 99.349% for pitch accents and 99.992% for endtones.

D.J.Hirst, A. Di Cristo & R. Espesser 2000 Levels of representation and levels of analysis for the description of intonation systems

Prosody: Theory and …, 2000

It is argued that a satisfactory global theory of intonation will require four levels of analysis : (i) physical (acoustic, physiological) (ii) phonetic (iii) surface phonological and (iv) deep phonological. The theoretical and cognitive status of each level is discussed and specific proposals are made for a model respecting such an overall architecture as well as a condition of interpretability which requires that each level of representation be interpretable in terms of adjacent levels. The level of phonetic representation is conceived of as providing an interface between abstract cognitive representations and their physical manifestations. This level is also assumed to provide an interface between constraints on production and perception. For fundamental frequency an algorithm, MOMEL, for the automatic derivation of a representation as a sequence of target-points is presented. The level of surface phonological representation is seen as the prosodic equivalent of the International Phonetic Alphabet for phonemic representation. A symbolic coding system for fundamental frequency patterns (INTSINT) is described which is currently being used for the automatic coding of fundamental frequency patterns for continuous texts in a number of different languages. The level of deep phonological representation is described as the level of linguistically significant choices which interact with a number of language-specific prosodic parameters to generate observed intonation patterns.

Generating German intonation with a trainable prosodic model

2006

A trainable prosodic model called SFC (Superposition of Functional Contours), proposed by Holm and Bailly, is here confronted to German intonation. Training material is the publicly available Siemens Synthesis Corpus that provides spoken utterances for high-quality speech synthesis. We describe the labeling framework and first evaluation results that compares the original prosody of test sentences of this corpus with their prosodic rendering by the proposed model and state-of-the-art systems available on-line on the web.

From information structure to intonation (original) (raw)

Related papers