US 5,384,892 A · Dynamic language model for speech recognition

Abstract

A method of speech recognition which determines acoustic features in a sound sample; recognizes words comprising the acoustic features based on a language model, which determines the possible sequences of words that may be recognized; and the selection of an appropriate response based on the words recognized. Information about what words may be recognized, under which conditions those words may be recognized, and what response is appropriate when the words are recognized, is stored, in a preferred embodiment, in a data structure called a speech rule. These speech rules are partitioned according to the context in which they are active. When speech is detected, concurrent with acoustic feature extraction, the current state of the computer system is used to determine which rules are active and how they are to be combined in order to generate a language model for word recognition. A language model is dynamically generated and used to find the best interpretation of the acoustic features as a word sequence. This word sequence is then matched against active rules in order to determine the appropriate response. Rules that match all or part of the word sequence contribute data structures representing the "meaning" of the word sequence, and these data structures are used by the rule actions in order to generate an appropriate response to the spoken utterance.

Claims7

Dependent claims

A method of speech recognition in a speech recognition system comprising the following steps:

a. determining acoustic features in a sound sample;
b. upon commencing said determination of said acoustic features, determining possible combinations of words which may be recognized by said speech recognition system and storing said possible combinations of words as a current language model, said current language model being generated from a plurality of speech rules each comprising a language model and an associated action, each said language model in each of said plurality of speech rules including a plurality of states, words defining transitions between said plurality of states, and terminal states;
c. upon the completion of said generation of said current language model, recognizing words comprising said acoustic features by traversing states in said current language model until reaching said terminal states in said current language model;
and
d. subsequent to said step of recognizing words, determining a matched speech rule from said plurality of speech rules used to create said current language model and said words and performing said action associated with said matched speech rule.

2 more elements

Dependencies are inferred from the claim text.

Dynamic language model for speech recognition

Life of this patent

Dynamic language model for speech recognition

Life of this patent

Abstract

Claims7