US4028496A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Digital speech detector
Assignee: AT&T Inc
•
Filing: 1976-08-17
•
Granted: 1977-06-07
•
Expires: 1996-08-17
Abstract
Abstract The detection sensitivity and noise rejection of an arrangement for detecting speech in the presence of noise is improved by accumulating the weighted differences between input signal samples and their short-term running average. The detector thus tracks ambient noise, providing an adaptive detection threshold such that detection sensitivity is increased in low noise environments without excessive false operation on high level noise. The peak average attained during an interval of speech is used to provide variable hangover upon cessation of speech, yielding greater hangover for weak talkers than for loud talkers. In an illustrative embodiment of the speech detector used in a speech interpolation system, protection is afforded also against false transmission path operation due to detection of speech echo.
Inventors
- 👤 Robert Ernest LaMarche
- 👤 Carl Jerome May, Jr.
- 👤 Timothy James Zebo
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-23 23:40
CPC Classifications
G; G10; G10L; G10L25/00; G10L25/78; H; H03; H03G; H03G3/00; H03G3/20; H03G3/30; H03G3/34; H03G3/342; H04; H04J; H04J3/00; H04J3/17; H04J3/175
Claims Analysis
Total Claims: 20
Claim 1
Independent Claim
65 words
A speech detector, means for receiving input signal samples, means for generating a representation of the average magnitude of said samples, means for comparing each said sample with said representation and for providing an output signal corresponding to the difference therebetween, and means for accumulating said output signals, said accumulating means generating different control signals corresponding to different predetermined levels of said accumulated output signals.
Claim 2
Dependent Claim
65 words
A speech detector, means for receiving input signal samples, means for generating a representation of the average magnitude of said samples, means for comparing each said sample with said representation and for providing an output signal corresponding to the difference therebetween, and means for accumulating said output signals, said accumulating means generating different control signals corresponding to different predetermined levels of said accumulated output signals.
Claim 3
Dependent Claim
30 words
A speech detector according to claim 1 wherein said accumulating means includes input means for weighting said output signals according to the corresponding differences between said samples and said representation.
Claim 4
Dependent Claim
33 words
A speech detector according to claim 1 wherein said comparing means includes means for quantizing said differences between said samples and said representation, and wherein said output signals correspond to said quantized differences.
Claim 5
Dependent Claim
15 words
A speech detector according to claim 1 wherein said receiving means includes highpass filter means.
Claim 6
Dependent Claim
48 words
A speech detector according to claim 1 further comprising output means responsive to a first predetermined one of said control signals for generating a speech activity indication and responsive to a second predetermined one of said control signals for terminating said indication after a fixed interval of time.
Claim 7
Dependent Claim
30 words
A speech detector according to claim 3 wherein said accumulating means includes input means for weighting said output signals according to the corresponding differences between said samples and said representation.
Claim 8
Dependent Claim
34 words
A speech detector according to claim 6 further comprising means for varying the rate of accumulation of at least one of said output signals in accordance with the peak average magnitude of said samples.
Claim 9
Dependent Claim
95 words
A speech detector, means including a highpass filter for receiving encoded input signal samples, averaging means for generating an encoded representation of the average magnitude of the filtered samples over a particular interval of time, comparator means for generating respective output signals corresponding to the quantized difference between each of said filtered samples and said encoded representation, means for accumulating said output signals and for generating different control signals corresponding to said accumulation exceeding different predetermined magnitudes, and output means responsive to a first one of said control signals for providing a speech activity indication.
Claim 10
Dependent Claim
95 words
A speech detector, means including a highpass filter for receiving encoded input signal samples, averaging means for generating an encoded representation of the average magnitude of the filtered samples over a particular interval of time, comparator means for generating respective output signals corresponding to the quantized difference between each of said filtered samples and said encoded representation, means for accumulating said output signals and for generating different control signals corresponding to said accumulation exceeding different predetermined magnitudes, and output means responsive to a first one of said control signals for providing a speech activity indication.
Claim 11
Dependent Claim
19 words
A speech detector according to claim 8 wherein said comparator output signals are weighted according to said quantized differences.
Claim 12
Dependent Claim
32 words
A speech detector according to claim 8 further comprising means responsive to a second one of said control signals for storing an encoded representation of the peak average magnitude of said samples.
Claim 13
Dependent Claim
63 words
A speech detector according to claim 10 further comprising timing means for generating a sequence of timing signals at a rate determined by said peak average magnitude representation and for extending said timing signals to said accumulating means, said accumulating means including means responsive to said timing signals and to a predetermined output signal generated by said comparator means for decrementing said accumulation.
Claim 14
Dependent Claim
26 words
A speech detector according to claim 11 further comprising means responsive to said accumulation falling below a certain magnitude for modifying said peak average magnitude representation.
Claim 15
Dependent Claim
28 words
A speech detector according to claim 11 wherein said accumulating means includes means for generating a third control signal corresponding to said accumulation falling below a fixed magnitude.
Claim 16
Dependent Claim
31 words
A speech detector according to claim 13 wherein said output means includes means responsive to said third control signal for terminating said speech activity indication after a fixed interval of time.
Claim 17
Dependent Claim
141 words
A speech detector, means for receiving input signal samples, means for generating a representation of the exponentially-mapped-past (EMP) average of said input samples, means for providing a first control signal when the magnitude of one of said input samples is less then two times said EMP average, means for providing a second control signal when the magnitude of one of said input samples is greater than four times said EMP average, means for providing a third control signal when the magnitude of one of said input samples is greater than two times said EMP average but is less than four times said EMP average, means for weighting each of said control signals differently, means for accumulating said weighted control signals, and means for providing a first output signal when the accumlation of said weighted control signals exceeds a predetermined threshold magnitude.
Claim 18
Dependent Claim
141 words
A speech detector, means for receiving input signal samples, means for generating a representation of the exponentially-mapped-past (EMP) average of said input samples, means for providing a first control signal when the magnitude of one of said input samples is less then two times said EMP average, means for providing a second control signal when the magnitude of one of said input samples is greater than four times said EMP average, means for providing a third control signal when the magnitude of one of said input samples is greater than two times said EMP average but is less than four times said EMP average, means for weighting each of said control signals differently, means for accumulating said weighted control signals, and means for providing a first output signal when the accumlation of said weighted control signals exceeds a predetermined threshold magnitude.
Claim 19
Dependent Claim
66 words
A speech detector according to claim 15 further comprising echo protection means including means for approximating echo anticipated in said input signal samples, means for providing an echo indication when the magnitude of one of said input signal samples is less than said echo approximation, and means responsive to said echo indication and said first output signal for controlling the transmission of said input signal samples.
Claim 20
Dependent Claim
29 words
A speech detector according to claim 16 wherein said controlling means includes means for disabling transmission of said input signal samples in the absence of said first output signal.
References Cited (Backward Citations)
US3520999A
Cited
US3644680A
Cited
US3712959A
Cited
US3649766A
Cited
US3801747A
Cited
US3832493A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1980-06-06 | 1980-12-06 | 1981-06-07 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1984-06-06 | 1984-12-06 | 1985-06-07 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 1988-06-06 | 1988-12-06 | 1989-06-07 | $7700.00 | Window Closed / Lapsed |