tailieunhanh - Báo cáo khoa học: "A Finite-Slate Parser for Use in Speech Recognition"

This paper is divided into two parts. 1 The first section motivates the application of finite-state parsing techniques at the phonetic level in order to exploit certain classes or" contextual constraints. -In the second section, the parsing framework is extended in order to account ['or 'feature spreading' (i:.g., agreement and co-articulation) in a natural way. | A Finite-State Parser for Use in Speech Recognition Kenneth w. Church NE43-307 Massachusetts Institute ofTcchnology Cambridge MA. 02139 This paper is divided into two Ihc first section motivates the application of finite-state parsing techniques at the phonetic level in order to exploit certain classes of contextual constraints. In the second section the parsing framework is extended in order to account for feature spreading . agreement and co-articulation in a natural way. 1. Parsing at the Phonetic Level It is well known that phonemes have different acoustic phonctic realizations depending on the context For example die phoneme ư is typically realized with a different allophone phonetic variant in syllable initial position than in syllabic final position. In syllabic initial position . Tom t is almost always released with a strong burst of energy and aspirated with h-likc noise whereas in syllable final position . cat zt is often unrelcascd and unaspirated. It is common practice in speech research to distinguish acoustic phonetic properties that vary a great deal with context . release and aspiration from those that are relatively invariant to context . place manner and voicing .2 In the past the emphasis has been on invariants allophonic variation is traditionally seen as problematic for recognition. 1 In most systems for sentence recognition such modifications must be viewed as a kind of noise that makes it more difficult to hypothesize lexical candidates given an input phonetic transcription. To see that this must be the case we note that each phonological rule in an example to be presented below 1. This research was supported in part by die National Institutes of I Icalth Grant No. 1 rot I M 03374-01 and 03374-02 from the National Library of .Medicine. 2. Place refers IO the location of Ute constriction in the vocal trace lixamples include labial al the lips Zp. b. f. V. mZ. velar Zk. g. T Z. dental at the teeth Zs. z. 1. d. I nZ and .