Investigation and modeling of coarticulation during speech
نویسندگان
چکیده
Coarticulation is an important phenomenon of speech production, which involves in both the physical level concerned with articulators’ properties, and the planning stage for generating motor commands. The authors have proposed a model for coarticulation, named “carrier model”, to imitate the coarticulation mechanism. The carrier model is built on an assumption that articulation can be separated into a vocalic movement and a consonantal movement, and that coarticulation takes place within and between the movements as a modulation process. This study first verifies the assumption using a reconstructed text-independent articulatory movement based on observed articulatory data. The carrier model is elaborated based on those two movements and the look-ahead mechanism is employed to materialize the coarticulation. Coefficients of the carrier model were quantified via statistical analysis of articulatory data. A numerical experiment was conducted by implementing the carrier model in a physiological articulatory model (Dang and Honda, 2004). The results showed that the carrier model gave a good performance for simulating coarticulation.
منابع مشابه
Allophone-based acoustic modeling for Persian phoneme recognition
Phoneme recognition is one of the fundamental phases of automatic speech recognition. Coarticulation which refers to the integration of sounds, is one of the important obstacles in phoneme recognition. In other words, each phone is influenced and changed by the characteristics of its neighbor phones, and coarticulation is responsible for most of these changes. The idea of modeling the effects o...
متن کاملVisual analysis of lip coarticulation in VCV utterances
This paper presents an investigation of the visual variation on the bilabial plosive consonant /p/ in three coarticulation contexts. The aim is to provide detailed ensemble analysis to assist coarticulation modelling in visual speech synthesis. The underlying dynamics of labeled visual speech units, represented as lip shape, from symmetric VCV utterances, is investigated. Variation in lip dynam...
متن کاملModeling between-word coarticulation in continuous speech recognition
This paper describes the addition of between-word coarticulation modeling into SPHINX, an accurate Iarge-vocabulary speakerindependent continuous speech recognition system. Between-word coarticulation is a major source of phonetic variability in continuous speech. By detailed modeling of between-word triphones and utilizing the generalized triphone technique, we obtain an error ;ate reduction o...
متن کاملModeling coarticulation in continuous speech
Modeling coarticulation in speech has been largely limited to short sequences and/or limited phonetic context. We introduce a methodology for modeling both formant frequency and bandwidth in continuous speech, allowing examination of sentencelevel coarticulation. The model represents continuous trajectories as a combination of overlapping local trajectories, which are represented by a weighted-...
متن کاملGestural overlap, place of articulation and speech rate - an x-ray investigation
This investigation deals with the production of consonant sequences in French, with particular focus on overlap of labial, apical, tongue-dorsum and velar gestures. X-ray and acoustic data are obtained for two speakers, at two speaking rates, normal-conversational and fast. Speech rate is varied in order to explore gestural overlapping possibilities, when places of articulation differ between p...
متن کامل