# US 11,451,905 B1 · System and method for multi-channel acoustic echo and feedback compensation > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,451,905 B1 - Document ID: US11451905B1 - Status: Granted - Kind: B1 - Application number: 17/086,295 - Filed: 2020-10-30 - Granted: 2022-09-20 - Earliest claimed: 2019-10-30 - Applicants: Social Microphone, Inc. - Inventors: David G. Shaw - Web page: https://patentgrep.com/patent/US11451905B1 ## CPC - G10L 21/0208: Noise filtering - H04R 3/02: for preventing acoustic reaction {, i.e. acoustic oscillatory feedback (specially adapted for hearing aids H04R25/453)} - G10L 21/02: Speech enhancement, e.g. noise reduction or echo cancellation (reducing echo effects in line transmission systems H04B3/20; echo suppression in hands-free telephones H04M9/08) - H04M 9/082: {using echo cancellers (echo cancellers per se H04B3/23)} - H04R 27/00: Public address systems (circuits for preventing acoustic reaction H04R3/02; circuits for distributing signals to loudspeakers H04R3/12; {monitoring or testing arrangements for public address systems H04R29/007}; amplifiers H03F) - H04S 7/305: {Electronic adaptation of stereophonic audio signals to reverberation of the listening space (H04S7/301 takes precedence)} - G10L 2021/02082: {the noise being echo, reverberation of the speech} - G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING - H04R: LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; ELECTRIC HEARING AIDS; PUBLIC ADDRESS SYSTEMS (producing sounds with frequency not determined by supply frequency G10K) - H04M: TELEPHONIC COMMUNICATION (circuits for controlling other apparatus via a telephone cable and not involving telephone switching apparatus G08) - H04S: STEREOPHONIC SYSTEMS ## Abstract A public address system includes audio inputs from a moderator/presenter and from one or more participants. In an embodiment, the moderator speaks first and then selects participants to speak, utilizing audio captured by participant devices. A central signal processor is configured to receive the audio inputs and to utilize a configured acoustic model to provide for acoustic echo cancellation (AEC) and feedback control (FBC) during various phases of a presentation or conference. Audio signals from the presenter and/or participants, that have been processed to remove echo, are utilized as reference signals during various phases of the audio presentation that utilize the acoustic model for either AEC or FBC. The system utilizes the knowledge that the best learning occurs during the far talking state to learn the echo path in the canceler mode (AEC) vs. the feedback mode (FBC), which usually can only train in a double-talking mode. ## Claims ### Claim 1 A method for reducing acoustic echo comprising: receiving by a central processing system, an acquired reference audio signal; receiving by the central processing system, an acquired presenter audio signal; receiving by the central processing system, an acquired at least one near end audio signal; preforming, by the central processing system acoustic model training of at least one near end audio signal based on audio input acquired by a signal processing unit, concurrent to processing and amplifying the presenter audio signal; performing, by the central processing system, echo cancellation of the at least one near end audio signal based on the corresponding at least one trained acoustic model. ### Claim 2 (depends on 1) The method for reducing acoustic echo as in claim 1 wherein the one or more near end audio signals is ascribed to a participant. ### Claim 3 (depends on 1) The method for reducing acoustic echo as in claim 1 wherein the one or more near end audio signal is ascribed to a participant and wherein the presenter audio signal is ascribed to a presenter. ### Claim 4 (depends on 3) The method for reducing acoustic echo as in claim 3 wherein the presenter speaks prior to a participant selection by the presenter from a selection queue, whereby performing acoustic model training of the at least one acoustic model prior to a participant speaking provides echo cancellation at the time the participant speaks. ### Claim 5 (depends on 1) The method for reducing acoustic echo as in claim 1 wherein the reference audio signal is prerecorded. ### Claim 6 (depends on 1) The method for reducing acoustic echo as in claim 1 wherein the reference audio signal is synthesized. ### Claim 7 (depends on 1) The method for reducing acoustic echo as in claim 1 wherein the reference audio signal comprises audio captured from a presentation audiovisual stream. ### Claim 8 (depends on 5) The method for reducing acoustic echo as in claim 5 also comprising: broadcasting the reference audio signal into an acoustic space prior to the step of performing acoustic model training. ### Claim 9 (depends on 1) The method for reducing acoustic echo as in claim 1 wherein the function of an acoustic canceler may be configured to switch from acoustic echo cancelling to feedback control by the act of having the presenter select a participant. ### Claim 10 (depends on 6) The method for reducing acoustic echo as in claim 6 wherein the acoustic canceler is configured to switch automatically from acoustic echo cancelling to feedback control when the participant's position in the queue is a next selection. ### Claim 11 (depends on 1) The method for reducing acoustic echo as in claim 1 wherein a presenter's feedback control is configured to adapt based on the presenter's microphone signal. ### Claim 12 (depends on 1) The method for reducing acoustic echo as in claim 1 wherein a selected participant's feedback control is configured to adapt based on the selected participant's processed microphone signal. ### Claim 13 (depends on 1) The method for reducing acoustic echo as in claim 1 wherein the echo cancellation is configurable to process the reference audio signal either for acoustic echo cancellation adaptation or for feedback control adaptation. ### Claim 14 (depends on 1) The method for reducing acoustic echo as in claim 1 wherein training of the acoustic model is configured to utilize as input a participant audio signal of one or more participants as an input parameter for learning an acoustic path of the respective participant. ### Claim 15 A system for acoustic echo cancelling comprising: a central processing system, wherein the central processing system receives an acquired reference audio signal, an acquired presenter audio signal and an acquired at least one near end audio signal; and wherein central processing system performs acoustic model training by at least one acoustic model of the at least one near end audio signal based on audio input acquired by a signal processing unit, concurrent to processing and amplifying the presenter audio signal; and wherein the central processing system performs echo cancellation of the at least one near end audio signal based on the corresponding at least one trained acoustic model. ### Claim 16 (depends on 15) The system for reducing acoustic echo as in claim 15 wherein the one or more near end audio signals is ascribed to a participant. ### Claim 17 (depends on 15) The system for reducing acoustic echo as in claim 15 wherein the one or more near end audio signal is ascribed to a participant and wherein the presenter audio signal is ascribed to a presenter. ### Claim 18 (depends on 17) The system for reducing acoustic echo as in claim 17 wherein the presenter speaks prior to a participant selection by the presenter from a selection queue, whereby performing acoustic model training of the at least one acoustic model prior to a participant speaking provides echo cancellation at the time the participant speaks. ### Claim 19 (depends on 15) The system for reducing acoustic echo as in claim 15 wherein the reference audio signal is prerecorded. ### Claim 20 (depends on 15) The system for reducing acoustic echo as in claim 15 wherein the reference audio signal is synthesized.