US5509104A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Speech recognition employing key word modeling and non-key word modeling
Assignee: AT&T Inc
•
Filing: 1993-10-06
•
Granted: 1996-04-16
•
Expires: 2013-10-06
Abstract
Abstract Speaker independent recognition of small vocabularies, spoken over the long distance telephone network, is achieved using two types of models, one type for defined vocabulary words (e.g., collect, calling-card, person, third-number and operator), and one type for extraneous input which ranges from non-speech sounds to groups of non-vocabulary words (e.g. `I want to make a collect call please`). For this type of key word spotting, modifications are made to a connected word speech recognition algorithm based on state-transitional (hidden Markov) models which allow it to recognize words from a pre-defined vocabulary list spoken in an unconstrained fashion. Statistical models of both the actual vocabulary words and the extraneous speech and background noises are created. A syntax-driven connected word recognition system is then used to find the best sequence of extraneous input and vocabulary word models for matching the actual input speech.
Inventors
- 👤 Chin H. Lee
- 👤 Lawrence R. Rabiner
- 👤 Jay G. Wilpon
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-29 06:57
CPC Classifications
G; G10; G10L; G10L15/00; G10L15/08; G10L15/14; G10L15/142; G10L2015/088
Claims Analysis
Total Claims: 24
Claim 1
Independent Claim
124 words
A method of processing an input signal representing a spoken utterance, the spoken utterance having a key utterance component and an extraneous sound component, the method comprising the steps of comparing the input signal to a plurality of speech recognition models within a speech recognition system, said plurality of speech recognition models including key word speech recognition models representative of respective different key utterances and further including at least a first sink model, and recognizing a particular one of said key utterances in said spoken utterance in response to said comparing, characterized in that said sink model represents a plurality of extraneous sound training tokens, at least two of said extraneous sound training tokens being other than repetitions of a particular one vocabulary item.
Claim 2
Independent Claim
124 words
A method of processing an input signal representing a spoken utterance, the spoken utterance having a key utterance component and an extraneous sound component, the method comprising the steps of comparing the input signal to a plurality of speech recognition models within a speech recognition system, said plurality of speech recognition models including key word speech recognition models representative of respective different key utterances and further including at least a first sink model, and recognizing a particular one of said key utterances in said spoken utterance in response to said comparing, characterized in that said sink model represents a plurality of extraneous sound training tokens, at least two of said extraneous sound training tokens being other than repetitions of a particular one vocabulary item.
Claim 3
Dependent Claim
29 words
The method of claim 1 wherein at least one of said two extraneous sound training tokens is a spoken utterance which is different from any of said key utterances.
Claim 4
Dependent Claim
22 words
The method of claim 1 wherein said extraneous sound training tokens include at least two of the utterances "um," "please," and "call."
Claim 5
Independent Claim
20 words
The method of claims 1, 2 or 3 wherein one of said extraneous sound training tokens is a background sound.
Claim 6
Dependent Claim
38 words
The method of claim 1 wherein individual tasks are associated with each of said key utterances and wherein said method comprises the further step of performing the task associated with the key utterance recognized in said recognizing step.
Claim 7
Dependent Claim
14 words
The method of claim 5 wherein said individual tasks are respective different operator-assisted-telephone-call tasks.
Claim 8
Dependent Claim
25 words
The method of claim 1 wherein in said speech recognition system, said plurality of speech recognition models are interrelated in accordance with a predefined grammar.
Claim 9
Dependent Claim
21 words
The method of claim 7 wherein said predefined grammar is a finite state grammar describing a set of valid spoken utterances.
Claim 10
Dependent Claim
25 words
The method of claim 1 wherein said speech recognition system implements a connected word speech recognition algorithm based on said plurality of speech recognition models.
Claim 11
Dependent Claim
31 words
The method of claim 1 wherein said speech recognition system is a grammar-driven continuous word recognition system in which the components of the grammar are represented by said speech recognition models.
Claim 12
Independent Claim
47 words
The method of claims 8, 9 or 10 wherein said grammar characterizes said speech input as an individual one of said key utterances, represented by said key word speech recognition models, preceded and/or succeeded by one or more extraneous sounds, represented by at least said sink model.
Claim 13
Dependent Claim
122 words
A speech recognition system for processing an input signal representing a spoken utterance, the spoken utterance having a key utterance component and an extraneous sound component, the speech recognition system comprising means for comparing the input signal to a plurality of speech recognition models, said plurality of speech recognition models including speech recognition models representative of respective different key utterances and further including at least a first sink model, and means for recognizing a particular one of said key utterances in said spoken utterance in response to said comparing, characterized in that said sink model represents a plurality of extraneous sound training tokens, at least two of said extraneous sound training tokens being other than repetitions of a particular one vocabulary item.
Claim 14
Dependent Claim
122 words
A speech recognition system for processing an input signal representing a spoken utterance, the spoken utterance having a key utterance component and an extraneous sound component, the speech recognition system comprising means for comparing the input signal to a plurality of speech recognition models, said plurality of speech recognition models including speech recognition models representative of respective different key utterances and further including at least a first sink model, and means for recognizing a particular one of said key utterances in said spoken utterance in response to said comparing, characterized in that said sink model represents a plurality of extraneous sound training tokens, at least two of said extraneous sound training tokens being other than repetitions of a particular one vocabulary item.
Claim 15
Dependent Claim
31 words
The speech recognition system of claim 12 wherein at least one of said two extraneous sound training tokens is a spoken utterance which is different from any of said key utterances.
Claim 16
Dependent Claim
24 words
The speech recognition system of claim 13 wherein said extraneous sound training tokens include at least two of the utterances "um," "please," and "call."
Claim 17
Dependent Claim
19 words
The speech recognition system of claim 13 wherein said plurality of extraneous sound training tokens includes a background sound.
Claim 18
Dependent Claim
32 words
The speech recognition system of any one of claims 12 through 15 wherein in said speech recognition system, said plurality of speech recognition models are interrelated in accordance with a predefined grammar.
Claim 19
Dependent Claim
18 words
The speech recognition system of claim 16 wherein said predefined grammar describes a set of expected spoken utterances.
Claim 20
Dependent Claim
46 words
The speech recognition system of claim 16 wherein said grammar characterizes said speech input as an individual one of said key utterances, represented by said key word speech recognition models, preceded and/or succeeded by one or more extraneous sounds, represented by at least said sink model.
Claim 21
Dependent Claim
38 words
The speech recognition system of any one of claims 12 through 15 wherein said speech recognition system is a grammar-driven continuous word recognition system in which the components of the grammar are represented by said speech recognition models.
Claim 22
Dependent Claim
32 words
The speech recognition system of any one of claims 12 through 15 wherein said speech recognition system implements a connected word speech recognition algorithm based on said plurality of speech recognition models.
Claim 23
Dependent Claim
31 words
The speech recognition system of claim 20 wherein said algorithm characterizes said speech input as an individual one of said key utterances preceded and/or succeeded by one or more extraneous sounds.
Claim 24
Dependent Claim
26 words
The invention of any one of claims 1, 2, 7, and 12 wherein each of said plurality of speech recognition models is a Hidden Markov Model.
References Cited (Backward Citations)
US4481593A
Cited
US4713777A
Cited
US5218668A
Cited
US4783804A
Cited
US4977599A
Cited
US4829577A
Cited
US4827521A
Cited
US4837831A
Cited
US4914703A
Cited
US5199077A
Cited
US5440662A
Cited
US5452397A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1999-04-16 | 1999-10-16 | 2000-04-16 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2003-04-16 | 2003-10-16 | 2004-04-16 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2007-04-16 | 2007-10-16 | 2008-04-16 | $7700.00 | Window Closed / Lapsed |