Proceedings ArticleDOI
Word verification in a speech understanding system
C. Cook
- Vol. 1, pp 553-556
Reads0
Chats0
TLDR
Verification offers an alternative strategy by doing a top-down parametric word match independent of segmentation and labeling, which results in a distance measure between the reference parameterization of a hypothesized word and the computed parameterizations of the real speech.Abstract:
If, in a speech understanding system, word matching is performed at the phonetic level, then the accurate determination of the locations and identities of words present in an unknown utterance is necessarily limited by the phonetic segmentation and labeling. Verification offers an alternative strategy by doing a top-down parametric word match independent of segmentation and labeling. The result is a distance measure between the reference parameterization of a hypothesized word and the computed parameterization of the real speech. This distance is interpreted as the likelihood of that word having actually occurred over a given portion of the utterance.read more
Citations
More filters
Proceedings ArticleDOI
The HWIM speech understanding system
TL;DR: The speech understanding system HWIM is designed to "understand" naturally spoken utterances relevant to a task domain of travel budget management and its performance results at the end of the project are presented.
Proceedings ArticleDOI
Time alignment of natural speech to synthetic speech
TL;DR: An artificial speech recognition experiment is introduced as a convenient means of assessing alignment accuracy, and alignment accuracy is found to be improved considerably by applying certain speaker adaptation transformations to the synthetic speech.
Journal ArticleDOI
Vowel normalization by frequency warped spectral matching
H Matsumoto,H Wakita +1 more
TL;DR: This paper presents an approach to vowel normalization based on frequency warped spectral matching, and an application to speaker normalization for word detection in connected speech is discussed.
Large Vocabulary Continuous Speech Recognition Using Linguistic Features and Constraints
TL;DR: A number of extensions to the original Huttenlocher-Zue lexical access model are made to take advantage of the existing facilities of a probabilistic, graph-based recognition framework and, more importantly, to model the broad linguistic features in a data-driven approach.
References
More filters
Journal ArticleDOI
Minimum prediction residual principle applied to speech recognition
TL;DR: A computer system is described in which isolated words, spoken by a designated talker, are recognized through calculation of a minimum prediction residual through optimally registering the reference LPC onto the input autocorrelation coefficients using the dynamic programming algorithm.
Linear Prediction and the Spectral Analysis of Speech
John Makhoul,Jared J. Wolf +1 more
TL;DR: In this article, a detailed treatment of the use of linear prediction in speech analysis is given, where covariance and autocorrelation methods are derived in the time and frequency domains.
Journal ArticleDOI
Structure of a phonological rule component for a synthesis-by-rule program
TL;DR: A synthesis-by-rule program can be thought of as a functional model of human sentence production that converts a linear string of symbols into a narrow phonetic transcription and a specification of stress levels, segmental durations, and aspects of the fundamental frequency contour.
Journal ArticleDOI
Structure of a phonological component for a synthesis‐by‐rule program
TL;DR: A synthesis-by-rule program can be thought of as a functional model of human sentence production as discussed by the authors, which accepts as input a linear string of symbols that have been produced by the semantics component, syntactic component, and lexical component of a grammar of English.