US4914692A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Automatic speech recognition using echo cancellation
Assignee: Nokia of America Corp
•
Filing: 1987-12-29
•
Granted: 1990-04-03
•
Expires: 2007-12-29
Abstract
Abstract A customer inquiry unit that allows the recognition of human speech in the presence of an initial prerecorded message transmitted by the customer inquiry unit via a telephone network. The customer inquiry system is designed to be utilized to respond to inquiries received via the telephone network. The system functions by responding to a telephone call received via the network to transmit an initial speech message to the customer. Due to a variety of network problems, a portion of the speech energy of the message is reflected back to the inquiry system. This reflected energy is commonly referred to as an echo. If the customer starts to speak before the inquiry system has finished the initial message, the system utilizes an echo canceler to remove the echo received back from the telephone network and only transfers the speech response to a recognition subsystem. The echo canceler adapts to the telephone network during the initial portion of the initial speech message before the customer can respond.
Inventors
- 👤 Walter T. Hartwell
- 👤 Mark A. Johnson
- 👤 Joseph Picone
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-30 21:34
CPC Classifications
H; H04; H04B; H04B3/00; H04B3/02; H04B3/20; H04B3/23; H04B3/238; H04M; H04M3/00; H04M3/42; H04M3/487; H04M3/493; G; G10; G10L; G10L21/00; G10L21/02; G10L21/0208; G10L2021/02082; H04M2201/00; H04M2201/40; H04M3/002
Claims Analysis
Total Claims: 8
Claim 1
Independent Claim
121 words
Apparatus for performing speech recognition in conjunction with a communication means and announcement means to decode a speech response received via said communication means to a speech message from said announcement means transmitted via said communication means, comprising: means for adapting to said communication means in response to said message and an echo of said transmitted message only during a predefined amount of time after initiation of said message to generate parameters; means responsive to said parameters for canceling said echo from the combination of the received speech response and said echo from said communication means to generate a modified signal; and means for recognizing said speech response from said modified signal to generate an encoded signal representing said speech response.
Claim 2
Dependent Claim
121 words
Apparatus for performing speech recognition in conjunction with a communication means and announcement means to decode a speech response received via said communication means to a speech message from said announcement means transmitted via said communication means, comprising: means for adapting to said communication means in response to said message and an echo of said transmitted message only during a predefined amount of time after initiation of said message to generate parameters; means responsive to said parameters for canceling said echo from the combination of the received speech response and said echo from said communication means to generate a modified signal; and means for recognizing said speech response from said modified signal to generate an encoded signal representing said speech response.
Claim 3
Dependent Claim
26 words
The apparatus of claim 1 wherein said predefined time is sufficient to enable said adapting means to adapt to the echo characteristics of said communication means.
Claim 4
Independent Claim
188 words
The apparatus for performing speech recognition in conjunction with a telephone network and a telephone to respond to inquiries made by a customer via said telephone and said telephone network, comprising: means responsive to the initiation of a telephone call via said network and said telephone to generate a first signal; means responsive to said first signal for communicating an initial speech message to said telephone via said network; means responsive to the initial transmission of said speech message for generating a second signal only during a predefined amount of time; means responsive to said second signal and said initial speech message and an echo of said communicated speech message received from said network for calculating parameters to adapt to conditions causing said echo within said network; means in response to said calculated parameters for removing said echo of said communicated speech message from the combination of the received speech response from said telephone set and said echo received from said telephone network to generate a modified signal; and means for recognizing said speech response from said modified signal to generate an encoded signal representing said speech response.
Claim 5
Independent Claim
188 words
The apparatus for performing speech recognition in conjunction with a telephone network and a telephone to respond to inquiries made by a customer via said telephone and said telephone network, comprising: means responsive to the initiation of a telephone call via said network and said telephone to generate a first signal; means responsive to said first signal for communicating an initial speech message to said telephone via said network; means responsive to the initial transmission of said speech message for generating a second signal only during a predefined amount of time; means responsive to said second signal and said initial speech message and an echo of said communicated speech message received from said network for calculating parameters to adapt to conditions causing said echo within said network; means in response to said calculated parameters for removing said echo of said communicated speech message from the combination of the received speech response from said telephone set and said echo received from said telephone network to generate a modified signal; and means for recognizing said speech response from said modified signal to generate an encoded signal representing said speech response.
Claim 6
Independent Claim
147 words
A method for performing speech recognition in an inquiry system have an announcement subsystem and a speech recognition subsystem in conjunction with a communication system for communicating the speech inquiries from a customer to said customer inquiry system, comprising the steps of: transmitting a speech message from said announcement subsystem to said customer via said communication system in response to initiation of one of said speech inquiries; adapting to said communication system in response to said transmitted message and an echo of said transmitted message only during a predefined time period to generate parameters; canceling in response to said parameters an echo of said transmitted message from the combination of the received speech response from said customer and said echo from said communication system to generate a modified signal; and recognizing said speech response from said modified signal to generate an encoded signal representing said speech response.
Claim 7
Independent Claim
147 words
A method for performing speech recognition in an inquiry system have an announcement subsystem and a speech recognition subsystem in conjunction with a communication system for communicating the speech inquiries from a customer to said customer inquiry system, comprising the steps of: transmitting a speech message from said announcement subsystem to said customer via said communication system in response to initiation of one of said speech inquiries; adapting to said communication system in response to said transmitted message and an echo of said transmitted message only during a predefined time period to generate parameters; canceling in response to said parameters an echo of said transmitted message from the combination of the received speech response from said customer and said echo from said communication system to generate a modified signal; and recognizing said speech response from said modified signal to generate an encoded signal representing said speech response.
Claim 8
Dependent Claim
26 words
The method of claim 4 wherein said predefined time is sufficient to enable said adapting means to adapt to the echo characteristics of said communication means.
References Cited (Backward Citations)
US4475189A
Cited
US4712235A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1993-04-02 | 1993-10-02 | 1994-04-03 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1997-04-02 | 1997-10-02 | 1998-04-03 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2001-04-02 | 2001-10-02 | 2002-04-03 | $7700.00 | Window Closed / Lapsed |