US3786188A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Synthesis of pure speech from a reverberant signal
Assignee: AT&T Inc
•
Filing: 1972-12-07
•
Granted: 1974-01-15
•
Expires: 1992-12-07
Abstract
Abstract Speech that has been reverberated by the transfer function of a reverberant enclosure is analyzed to detect parameters from which an unreverberative synthetic version of the original speech may be constructed. The process involves continuously approximating the vocal tract transfer function of the speaker. The effect of this transfer function is then removed from the reverberant speech by inverse filtering, the residual signal being the glottis excitation signal reverberated by the room. The reverberant excitation function is then analyzed to determine when the speaker''s driving function is voiced or unvoiced, the periodicity when voiced, and a unique gain factor. Then clean speech is synthesized using the foregoing three parameters operating on an all-pole filter that is continuously adapted to approximate the vocal tract transfer function.
Inventors
- 👤 J Allen
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-25 01:14
CPC Classifications
G; G10; G10L; G10L25/00; G10K; G10K2210/00; G10K2210/30; G10K2210/301; G10K2210/3013; G10K2210/50; G10K2210/505; G10L21/00; G10L21/02; G10L21/0208; G10L2021/02082
Claims Analysis
Total Claims: 13
Claim 1
Independent Claim
96 words
Apparatus for synthesizing speech comprising: transducer means located within a reverberant enclosure remotely from a speaker therein, for receiving reverberated speech signals from said speaker; means for continuously deriving, from said reverberated speech signals, first signals representative of the vocal tract transfer function of said speaker; means for developing, from said reverberated speech signals and said first signals, second signals representing the reverberated excitation source of said speaker; means for dereverberating said second signals; and means for developing, from said first signals and said dereverberated second signals, synthetic speech signals substantially approximating said speaker''s original speech.
Claim 2
Dependent Claim
96 words
Apparatus for synthesizing speech comprising: transducer means located within a reverberant enclosure remotely from a speaker therein, for receiving reverberated speech signals from said speaker; means for continuously deriving, from said reverberated speech signals, first signals representative of the vocal tract transfer function of said speaker; means for developing, from said reverberated speech signals and said first signals, second signals representing the reverberated excitation source of said speaker; means for dereverberating said second signals; and means for developing, from said first signals and said dereverberated second signals, synthetic speech signals substantially approximating said speaker''s original speech.
Claim 3
Dependent Claim
131 words
Apparatus for constructing an undistorted replica of a speaker''s original speech uttered in a reverberant enclosure comprising means for continuously extracting from the reverberant speech signal an approximation of the vocal tract transfer function of said speaker; means for removing from said reverberant speech the effect of said vocal tract transfer function, the resulting residual signal being substantially the speaker''s reverberated excitation function; means for deriving, from said reverberated excitation function, a first parameter denoting the voiced or unvoiced nature of said excitation function; a second parameter denoting the pitch period of voiced portions of said excitation function; and a third parameter denoting the mean-squared level of said excitation function; and means for combining said first, second, and third parameters with said vocal tract transfer function to produce said undistorted replica.
Claim 4
Dependent Claim
131 words
Apparatus for constructing an undistorted replica of a speaker''s original speech uttered in a reverberant enclosure comprising means for continuously extracting from the reverberant speech signal an approximation of the vocal tract transfer function of said speaker; means for removing from said reverberant speech the effect of said vocal tract transfer function, the resulting residual signal being substantially the speaker''s reverberated excitation function; means for deriving, from said reverberated excitation function, a first parameter denoting the voiced or unvoiced nature of said excitation function; a second parameter denoting the pitch period of voiced portions of said excitation function; and a third parameter denoting the mean-squared level of said excitation function; and means for combining said first, second, and third parameters with said vocal tract transfer function to produce said undistorted replica.
Claim 5
Dependent Claim
43 words
Apparatus pursuant to claim 2 wherein said extracting means further comprises means for recurrently estimating a sequence of weighting coefficients ak which constitute a unique estimate of a time-varying filter that approximates in its transfer function T( omega ) said speaker''s vocal tract.
Claim 6
Dependent Claim
26 words
Apparatus pursuant to claim 3 wherein said removing means comprises an inverse filter having as its inputs said weighting coefficients ak and said reverberant speech signal.
Claim 7
Dependent Claim
45 words
Apparatus pursuant to claim 4 wherein said means for deriving said second parameter comprises means for autocorrelating said excitation function to determine a maximum value, and means for ascertaining a unique delay associated with said maximum value, said delay constituting the said pitch period parameter.
Claim 8
Dependent Claim
58 words
Apparatus pursuant to claim 5 wherein said means for deriving said first parameter comprises a fixed threshold detector for inspecting the level of each of said maximum values resulting from said autocorrelation of said excitation function, and means for producing a voiced-unvoiced decision based on whether a given said maximum value exceeds or falls below said fixed threshold.
Claim 9
Dependent Claim
68 words
Apparatus pursuant to claim 6 further comprising: a white noise source, a variable period pulse generator, means for transmitting said pitch period parameter to said pulse generator to control the period of said pulses, and a voicing switch responsive to the output of said threshold detector for selecting as its output either said white noise source or said pulse generator, thereby to produce a fixed amplitude source signal.
Claim 10
Dependent Claim
51 words
Apparatus pursuant to claim 7, wherein said means for generating said third parameter comprises means for multiplying said fixed amplitude source signal by an amount that renders the speaker''s said reverberated excitation function identical in mean-squared level with said fixed amplitude source signal, the result being a synthetic unreverberated excitation function.
Claim 11
Dependent Claim
42 words
Apparatus pursuant to claim 8, wherein said combining means comprises an all-pole filter having as its inputs said weighting coefficients ak and said unreverberated excitation function, the output of said combining means constituting the speaker''s synthesized speech signal free of reverberative effects.
Claim 12
Dependent Claim
148 words
A speech dereverberation system for an enclosure characterized by a fixed transfer function H( omega ), comprising: a source of white noise; a source of electrical impulses; a recursive filter having a variable transfer function; means for receiving a reverberative speech signal generated from within said enclosure; means for deriving, from successive discrete sample sets of said speech signal, a specific said filter setting representing a currently valid vocal tract transfer function; means for deriving an indicia of the pitch period and an indicia of the ratio of voiced-to-unvoiced intervals in said speech signal; means for applying said pitch period indicia to said impulse source to control the impulse generation rate; a voiced-unvoiced interval ratio detector having a fixed threshold level; and means including said detector for applying said impulses to said recursive filter when said threshold is exceeded and otherwise for applying said noise to said filter.
Claim 13
Dependent Claim
148 words
A speech dereverberation system for an enclosure characterized by a fixed transfer function H( omega ), comprising: a source of white noise; a source of electrical impulses; a recursive filter having a variable transfer function; means for receiving a reverberative speech signal generated from within said enclosure; means for deriving, from successive discrete sample sets of said speech signal, a specific said filter setting representing a currently valid vocal tract transfer function; means for deriving an indicia of the pitch period and an indicia of the ratio of voiced-to-unvoiced intervals in said speech signal; means for applying said pitch period indicia to said impulse source to control the impulse generation rate; a voiced-unvoiced interval ratio detector having a fixed threshold level; and means including said detector for applying said impulses to said recursive filter when said threshold is exceeded and otherwise for applying said noise to said filter.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1977-01-14 | 1977-07-16 | 1978-01-15 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1981-01-14 | 1981-07-16 | 1982-01-15 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 1985-01-14 | 1985-07-16 | 1986-01-15 | $7700.00 | Window Closed / Lapsed |