System for distinguishing or counting spoken itemized expressions

Claim 2

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, a method for recognizing a spoken itemized expression speech comprising the following steps: sensing said waveform; identifying at least one high amplitude segment and at least one low amplitude segment in said sensed waveform, said low amplitude segment being adjacent and subsequent in time to said high amplitude segment; measuring a duration of said identified high amplitude segment with respect to said identified low amplitude segment; and producing a signal indicating a spoken itemized expression has been recognized when said high amplitude segment is shorter in duration than a threshold value, said threshold value being no greater than 20 times the duration of said low amplitude segment.

Claim 3

A method as claimed in claim 1 wherein the step of producing a signal indicating a spoken itemized expression has been recognized includes the step of determining a ratio indicative of comparative duration of said identified high and identified low amplitude segments and producing a signal indicating a spoken itemized expression has been recognized when said ratio is within a selected range of values.

Claim 4

A method as claimed in claim 1 wherein said spoken itemized expression which is to be recognized comprises a character.

Claim 5

A method as claimed in claim 1 wherein said threshold value is substantially no greater than twice said duration of said adjacent and low amplitude segment.

Claim 6

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, a method for recognizing a spoken itemized expression comprising the following steps: sensing said waveform; identifying at least one high amplitude segment and at least one low amplitude segment in said sensed waveform, said low amplitude segment being adjacent and subsequent in time to said high amplitude segment; measuring a duration of said identified high amplitude segment prior in time to said identified low amplitude segment; and producing a signal indicating a spoken itemized expression has been recognized when said high amplitude segment has a duration within a selected range of values.

Claim 7

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, a method for recognizing a spoken itemized expression comprising the following steps: sensing said waveform; identifying at least one high amplitude segment and at least one low amplitude segment in said sensed waveform, said low amplitude segment being adjacent and subsequent in time to said high amplitude segment; measuring a duration of said identified high amplitude segment prior in time to said identified low amplitude segment; and producing a signal indicating a spoken itemized expression has been recognized when said high amplitude segment has a duration within a selected range of values.

Claim 8

A method as claimed in claim 5 wherein said selected range of values is greater than 10 milliseconds and less than 250 milliseconds.

Claim 9

A method as claimed in claim 5 wherein said selected range of values includes values less than a specified maximum value.

Claim 10

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, a method for recognizing a spoken itemized expression comprising the following steps: sensing said waveform; identifying at least one high amplitude segment and at least one low amplitude segment in said sensed waveform, said low amplitude segment being adjacent and subsequent in time to said high amplitude segment; measuring a duration of said identified high amplitude segment with respect to said identified low amplitude segment; determining a ratio indicative of comparative duration of said identified high and identified low amplitude segments; and producing a signal indicating a spoken itemized expression has been recognized when said ratio is within a selected range of values.

Claim 11

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, a method for recognizing a spoken itemized expression comprising the following steps: sensing said waveform; identifying at least one high amplitude segment and at least one low amplitude segment in said sensed waveform, said low amplitude segment being adjacent and subsequent in time to said high amplitude segment; measuring a duration of said identified high amplitude segment with respect to said identified low amplitude segment; determining a ratio indicative of comparative duration of said identified high and identified low amplitude segments; and producing a signal indicating a spoken itemized expression has been recognized when said ratio is within a selected range of values.

Claim 12

A method as claimed in claim 1 or 5 or 8 wherein said itemized expression which is to be recognized is selected from the group consisting of individual words, digits, and characters.

Claim 13

A method as claimed in claim 1 or 5 or 8 wherein said identified high and low amplitude segments each comprise a plurality of segments and wherein said method further comprises the step of counting each high amplitude segment as a spoken itemized expression.

Claim 14

A method as claimed in claim 10 wherein said step of counting each high amplitude segment comprises the step of interrupting counting when a selected number of high amplitude segments are counted as spoken itemized expressions.

Claim 15

A method as claimed in claim 10 wherein said step of counting each high amplitude segment comprises the step of interrupting counting after a selected counting interval has elapsed.

Claim 16

A method as claimed in claim 10 wherein the step of counting each high amplitude segment comprises the step of interrupting counting in response to a high amplitude segment which is substantially longer in duration than an adjacent and subsequent in time low amplitude segment.

Claim 17

A method as claimed in claim 10 wherein the counting step comprises the step of interrupting counting in response to a high amplitude segment having a duration longer than a selected value.

Claim 18

A method as claimed in claim 14 wherein said selected value varies as a function of said duration of said adjacent low amplitude segment.

Claim 19

A method as claimed in claim 1 or 5 or 8 further comprising the step of recognizing a transition from a spoken itemized expression and responsively producing a transition signal.

Claim 20

A method as claimed in claim 16 wherein the step of producing a transition signal comprises the step of producing a signal in response to a first spoken itemized expression recognized.

Claim 21

A method as claimed in claim 16 wherein the step of producing a transition signal comprises the step of producing a signal in response to a last spoken itemized expression recognized.

Claim 22

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, said waveform comprising at least one high amplitude segment and at least one low amplitude segment, at least some of said high amplitude segments indicating speech, at least some of said low amplitude segments indicating periods of non-speech between adjacent speech segments, a speech categorization system for counting spoken itemized expressions comprising, in combination: sensing means for sensing said waveform; timer means, interconnected to said sensing means, for generating first and second measured durations, wherein said first measured durations vary as a function of duration of high amplitude segments, and wherein said second measured durations vary as a function of low amplitude segments; classification means for classifying a high amplitude segment in a first category when said first and second measured durations corresponding to said high amplitude segment and an adjacent and subsequent in time low amplitude segment satisfy a classification test; and a counter for counting said high amplitude segments classified in said first category, said high amplitude segments classified in said first category corresponding to said spoken itemized expressions.

Claim 23

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, said waveform comprising at least one high amplitude segment and at least one low amplitude segment, at least some of said high amplitude segments indicating speech, at least some of said low amplitude segments indicating periods of non-speech between adjacent speech segments, a speech categorization system for counting spoken itemized expressions comprising, in combination: sensing means for sensing said waveform; timer means, interconnected to said sensing means, for generating first and second measured durations, wherein said first measured durations vary as a function of duration of high amplitude segments, and wherein said second measured durations vary as a function of low amplitude segments; classification means for classifying a high amplitude segment in a first category when said first and second measured durations corresponding to said high amplitude segment and an adjacent and subsequent in time low amplitude segment satisfy a classification test; and a counter for counting said high amplitude segments classified in said first category, said high amplitude segments classified in said first category corresponding to said spoken itemized expressions.

Claim 24

A system as claimed in claim 19 wherein said classification means comprises: determination of means for establishing a ratio of said first and second durations; and comparison means for comparing said ratio to a selected value and for classifying said high amplitude segment in said first category when said ratio has a selected relationship to said selected value.

Claim 25

In a system having a voice processor for processing speech and responsively producing a waveform representing said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, said waveform comprising at least one high amplitude segment and at least one low amplitude segment, at least some of said high amplitude segments indicating speech, at least some of said low amplitude segments indicating periods of non-speech between adjacent speech segments, a speech categorization system for counting spoken itemized expressions comprising, in combination: sensing means for sensing said waveform; a timer, interconnected to said sensing means, for generating first measured durations, wherein said first measured durations vary as a function of duration of high amplitude segments; specification means for classifying a high amplitude segment in a first category, when the respective first measured duration satisfies a classification test; and a counter for counting high amplitude segments classified in said first category, said high amplitude segments classified in said first category corresponding to said spoken itemized expressions.

Claim 26

In a system having a voice processor for processing speech and responsively producing a waveform representing said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, said waveform comprising at least one high amplitude segment and at least one low amplitude segment, at least some of said high amplitude segments indicating speech, at least some of said low amplitude segments indicating periods of non-speech between adjacent speech segments, a speech categorization system for counting spoken itemized expressions comprising, in combination: sensing means for sensing said waveform; a timer, interconnected to said sensing means, for generating first measured durations, wherein said first measured durations vary as a function of duration of high amplitude segments; specification means for classifying a high amplitude segment in a first category, when the respective first measured duration satisfies a classification test; and a counter for counting high amplitude segments classified in said first category, said high amplitude segments classified in said first category corresponding to said spoken itemized expressions.

Claim 27

A system as claimed in claim 19 or 21 wherein said first category is indicative of a voiced expression.

Claim 28

A system as claimed in claim 19 or 21 further comprising interruption means for interrupting operation of said counter after a selected interval.

Claim 29

A system as claimed in claim 19 or 21 further comprising interruption means for interrupting operation of said counter after a selected number of high amplitude segments have been counted.

Claim 30

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, said waveform comprising at least one high amplitude segment and at least one low amplitude segment, at least some of said high amplitude segments indicating speech, at least some of said low amplitude segments indicating periods of non-speech between adjacent speech segments, a speech categorization system for classifying spoken itemized expressions comprising, in combination: sensing means for sensing said waveform; timer means, interconnected to said sensing means, for generating first and second measured durations, wherein said first measured durations vary as a function of duration of high amplitude segments, and wherein said second measured durations vary as a function of low amplitude segments; and classification means for classifying a high amplitude segment as a spoken itemized expression based on when said first and second measured durations for said higher amplitude segment and said adjacent said subsequent low amplitude segment satisfy a classification test.

Claim 31

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, said waveform comprising at least one high amplitude segment and at least one low amplitude segment, at least some of said high amplitude segments indicating speech, at least some of said low amplitude segments indicating periods of non-speech between adjacent speech segments, a speech categorization system for classifying spoken itemized expressions comprising, in combination: sensing means for sensing said waveform; timer means, interconnected to said sensing means, for generating first and second measured durations, wherein said first measured durations vary as a function of duration of high amplitude segments, and wherein said second measured durations vary as a function of low amplitude segments; and classification means for classifying a high amplitude segment as a spoken itemized expression based on when said first and second measured durations for said higher amplitude segment and said adjacent said subsequent low amplitude segment satisfy a classification test.

Claim 32

A system as claimed in claim 25 wherein said classification means comprises: determination means for taking a ratio of said first and second durations; and comparison means for comparing said ratio to a selected value and for classifying said high amplitude segment as a spoken itemized expression when said ratio has a selected relationship to said selected value.

Claim 33

A system as claimed in claim 25 wherein said classification means comprises comparison means for comparing said first and second durations and for classifying said high amplitude segment as a spoken itemized expression when said first duration is less than said second duration.

Claim 34

A system as claimed in claim 25 further comprising a counter for counting said high amplitude segments classified as spoken itemized expressions.

Claim 35

A system as claimed in claim 28 further comprising interruption means for interrupting operation of said counter after a selected interval.

Claim 36

A system as claimed in claim 28 further comprising interruption means for interrupting operation of said counter after a selected number of high amplitude segments have been counted.

Claim 37

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, said waveform comprising at least one high amplitude segment and at least one low amplitude segment, at least some of said high amplitude segments indicating speech, at least some of said low amplitude segments indicating periods of non-speech between adjacent speech segments, speech categorization system for counting spoken itemized expressions comprising in combination: parameter means for developing a classification parameter indicative of comparative duration of at least some of said higher amplitude segments with respect to adjacent low amplitude segments; classification means for classifying said high amplitude segments into at least two categories, wherein one of said categories is indicative of a spoken itemized expression, said classification means applying at least one classification test, said test placing one of said high amplitude segments in a first category when the respective classification parameter meets a threshold value and placing one of the said high amplitude segments in said second category when the respective classification parameter fails to meet said threshold value; and counting means for counting said high amplitude segments in said first category to characterize said voice message.

Claim 38

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, said waveform comprising at least one high amplitude segment and at least one low amplitude segment, at least some of said high amplitude segments indicating speech, at least some of said low amplitude segments indicating periods of non-speech between adjacent speech segments, speech categorization system for counting spoken itemized expressions comprising in combination: parameter means for developing a classification parameter indicative of comparative duration of at least some of said higher amplitude segments with respect to adjacent low amplitude segments; classification means for classifying said high amplitude segments into at least two categories, wherein one of said categories is indicative of a spoken itemized expression, said classification means applying at least one classification test, said test placing one of said high amplitude segments in a first category when the respective classification parameter meets a threshold value and placing one of the said high amplitude segments in said second category when the respective classification parameter fails to meet said threshold value; and counting means for counting said high amplitude segments in said first category to characterize said voice message.

Claim 39

A system as claimed in claim 31 wherein each of said classification parameter is indicative of a ratio between said duration of one of said high amplitude segments and said duration of an adjacent one of said low amplitude segments.

Claim 40

A system as claimed in claim 31 wherein said parameter means comprises: a first timer for timing duration of at least some high amplitude segments to generate a first measured duration; a second timer for timing duration of at least some low amplitude segments adjacent to said selected high amplitude segments to generate a second measured duration; and determination means for taking a ratio of said first and second measured durations, wherein said classification parameter varies as a function of said ratio.

Claim 41

A system as claimed in claim 31 wherein said first category is indicative of a spoken itemized expression and wherein said system further comprises means for classifying a high amplitude segment in said second category when said high amplitude segment has a duration greater than a selected value.

Claim 42

A system as claimed in claim 31 further comprising interruption means for interrupting operation of said counting means after a selected interval.

Claim 43

A system as claimed in claim 31 further comprising interruption means for interrupting operation of said counting means after a selected number of high amplitude segments have been counted.

Claim 44

A system as claimed in claim 31 wherein said spoken itemized expression is a voice character.

Claim 45

In a system having a voice processor for processing speech and responsively producing a waveform representing said information of speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, a method for recognizing a spoken itemized expression comprising the following steps: sensing said waveform; identifying at least one high amplitude segment and at least one low amplitude segment in said sensed waveform, said low amplitude segment being adjacent and subsequent in time to said high amplitude segment; measuring a duration of said identified high amplitude segment with respect to said identified low amplitude segment; producing an alternate signal, in an absence of a spoken itemized expression, when said high amplitude segment is substantially longer in duration than said adjacent and subsequent in time low amplitude segment.

Claim 46

In a system having a voice processor for processing speech and responsively producing a waveform representing said information of speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, a method for recognizing a spoken itemized expression comprising the following steps: sensing said waveform; identifying at least one high amplitude segment and at least one low amplitude segment in said sensed waveform, said low amplitude segment being adjacent and subsequent in time to said high amplitude segment; measuring a duration of said identified high amplitude segment with respect to said identified low amplitude segment; producing an alternate signal, in an absence of a spoken itemized expression, when said high amplitude segment is substantially longer in duration than said adjacent and subsequent in time low amplitude segment.

Claim 47

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, a method for recognizing a spoken itemized expression comprising the following steps: sensing said waveform; identifying at least one high amplitude segment and at least one low amplitude segment in said sensed waveform, said low amplitude segment being adjacent and subsequent in time to said high amplitude segment; measuring a duration of said identified high amplitude segment prior in time to said identified low amplitude segment; producing an alternate signal, in an absence of a spoken itemized expression, when said high amplitude segment has a measured duration within a selected range of values.

Claim 48

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, a method for recognizing a spoken itemized expression comprising the following steps: sensing said waveform; identifying at least one high amplitude segment and at least one low amplitude segment in said sensed waveform, said low amplitude segment being adjacent and subsequent in time to said high amplitude segment; measuring a duration of said identified high amplitude segment prior in time to said identified low amplitude segment; producing an alternate signal, in an absence of a spoken itemized expression, when said high amplitude segment has a measured duration within a selected range of values.

Claim 49

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, a method for recognizing a spoken itemized expression comprising the following steps: sensing said waveform; identifying at least one high amplitude segment and at least one low amplitude segment in said sensed waveform, said low amplitude segment being adjacent and subsequent in time to said high amplitude segment; measuring a duration of said identified high amplitude segment with respect to said identified low amplitude segment; determining a ratio indicative of comparative duration of said high and low amplitude segments; and producing an alternate signal in an absence of a spoken itemized expression, when said ratio is within a preselected range of values.

Claim 50

In a system having a voice processor for processing speech and responsively producing a waveform representing information of said speech, said waveform defining a high amplitude segment above a predetermined reference level and a low amplitude segment below said predetermined reference level, a method for recognizing a spoken itemized expression comprising the following steps: sensing said waveform; identifying at least one high amplitude segment and at least one low amplitude segment in said sensed waveform, said low amplitude segment being adjacent and subsequent in time to said high amplitude segment; measuring a duration of said identified high amplitude segment with respect to said identified low amplitude segment; determining a ratio indicative of comparative duration of said high and low amplitude segments; and producing an alternate signal in an absence of a spoken itemized expression, when said ratio is within a preselected range of values.

Claim 51

A method as claimed in claim 38 or 39 or 40 wherein said spoken itemized expression is selected from the group consisting of words, digits, and characters.