A hierarchical intonation model for synthesising F0 contours in galician language
نویسندگان
چکیده
In this contribution we propose a hierarchical intonation model for synthesising f0 contours with application to text-to-speech synthesis in Galician language. This model makes use of the implicit knowledge that resides in a database of natural f0 contours obtained from a read corpus. The novelty of this method lies on the way the f0 contour is generated. First, no phonological description in terms of a sequence of tones is needed prior to f0 generation. The phrasing obtained from previous stages of the TTS system is enough for this task. Second, the final f0 contour is built through several steps that assign patterns at the phonic group level (intonational phrase), the tonic group level and the segmental level following a hierarchical method. The proposed algorithm guarantees a coherent concatenation of the patterns that belong to different levels, and it seems to work properly as a general intonation model for a wide range of sentence modalities.
منابع مشابه
Modeling DCT parameterized F0 trajectory at intonation phrase level with DNN or decision tree
In the conventional HMM-based TTS, the micro structure of F0 contour is modeled at the state level via a (clustered) decision tree. However, the decision tree based state-level modeling is difficult to capture the long term structure of speech prosody, say at intonation phrase level, due to its greedy search nature and usually sparse training data for covering a large, combinatorial number of u...
متن کاملModeling segment intonation for Slovene TTS system
A scheme for modeling the F0 contour for different types of intonation units for the Slovene language is presented. It is based on results of analyzing F0 contours, using a quantitative model. Data from ten speakers was collected, resulting in a large corpora, mainly of declarative sentences. A way of generating the F0 contour for given utterances was defined, using only the text of the utteran...
متن کاملSynthesizing intonation of standard arabic language
In this paper, we propose a model to generate fundamental frequency (F0) contours using neural networks. A learning procedure is proposed as an alternative to synthesis-by-rules. The generation of correct fundamental frequency contour is one of the important issues in the naturalness of automatic text-to-speech conversion systems. The proposed approach is based on a standard feed-forward multi-...
متن کاملTransmitting Tone and Intonation Simultaneously — The Parallel Encoding and Target Approximation (PENTA) Model
Lexical tones use F0 to distinguish between words that are otherwise phonemically identical. Intonation uses F0 to convey discourse, attitudinal and affective information that is often not directly encoded in the words or syntax of the spoken utterances. Because the same acoustic parameter is being used, it is a question how well lexical tones and intonation can coexist in a language. The Paral...
متن کاملIntonation modelling with a lexicon of natural F0 contours
We describe a new approach for generating Norwegian intonation in text to speech synthesis. The method is based on a phonological representation of utterances. The overall f0 contour of an utterance is synthesised by concatenation of stored f0 contours corresponding to accent units. Candidate accent units are found by searching a lexicon derived from natural speech and selecting the unit that i...
متن کامل