Back to results
US 9,554,210 B1Granted
Multichannel acoustic echo cancellation with unique individual channel estimations
AssigneeAmazon Technologies, Inc.
InventorsRobert Ayrapetian, Michael Alan Pogue, Kenneth Edgar Hild, II
Life of this patent
Estimating the term…
Abstract
A multi-channel echo cancellation system that dynamically adapts to changes in acoustic conditions. The system does not require a sequence of “start-up” tones to determine the impulse responses. Rather, the adaptive filters approximate estimated transfer functions for each channel. A secondary adaptive filter adjusts cancellation to adapt to changes in the actual transfer functions over time after the adaptive filters have been trained, even if the reference signals are not unique relative to each other.
Claims20
A method, comprising:
- receiving a first signal including first audio and a second signal including second audio;
- transmitting the first signal to a first speaker, the first speaker outputting the first audio;
- transmitting the second signal to a second speaker, the second speaker outputting the second audio;
- receiving a detected signal from a microphone, the detected signal including the first audio output by the first speaker and the second audio output by the second speaker;
- applying first filter coefficients of a first adaptive filter to the first signal to generate a first estimated signal;
- causing the first adaptive filter to converge the first filter coefficients;
- applying second filter coefficients of a second adaptive filter to the second signal to generate a second estimated signal;
- causing the second adaptive filter to converge the second filter coefficients;
- determining a sum of the first estimated signal and the second estimated signal;
- subtracting the sum from the detected signal to determine a first output signal;
- determining that that the first filter coefficients of the first adaptive filter and the second filter coefficients of the second adaptive filter have converged to approximately stable values;
- in response to determining that the first and second adaptive filters have converged:
- fixing the first and second filter coefficients at the approximately stable values;
- applying third filter coefficients of a third adaptive filter to the sum to generate an estimated signal;
- causing the third adaptive filter to converge the third filter coefficients;
- and
- subtracting the estimated signal from the detected signal to determine a second output signal.
14 more elements
A computing device comprising:
- a processor;
- a memory including instruction operable to be executed by the processor to perform a set of actions to configure the processor to:
- receive a first audio signal and a second audio signal;
- transmit a first audio signal to a first sound reproduction device and the second audio signal to a second sound reproduction device;
- receive a detected signal from a microphone including a first portion of the first audio signal and a second portion of the second audio signal;
- provide the first audio signal as input to a first adaptive filter to generate a first estimated signal output, and the second audio signal as input to a second adaptive filter to generate a second estimated signal output;
- determine a sum of the first estimated signal output and the second estimated signal output;
- subtract the sum from the detected signal to determine a first output signal;
- adapt first filter coefficients of the first adaptive filter and second filter coefficients of the second adaptive filter to converge;
- determine that the first adaptive filter and the second adaptive filter have converged;
- disable adapting of the first and second filter coefficients after the first and second adaptive filters are determined to have converged;
- provide the sum as input to a third adaptive filter to generate a third estimated signal output;
- subtract the third estimated signal from the detected signal to determine a second output signal;
- and
- adapt third filter coefficients of the third adaptive filter to converge.
12 more elements
A method to be performed by a device, the method comprising:
- receiving a first audio signal and a second audio signal;
- transmitting a first audio signal to a first sound reproduction device and the second audio signal to a second sound reproduction device;
- receiving a detected signal from a microphone including a first portion of the first audio signal and a second portion of the second audio signal;
- providing the first audio signal as input to a first adaptive filter to generate a first estimated signal output, and the second audio signal as input to a second adaptive filter to generate a second estimated signal output;
- determining a sum of the first estimated signal output and the second estimated signal output;
- subtracting the sum from the detected signal to determine a first output signal;
- adapting first filter coefficients of the first adaptive filter and second filter coefficients of the second adaptive filter to converge;
- determining that the first adaptive filter and the second adaptive filter have converged;
- disabling adapting of the first and second filter coefficients after the first and second adaptive filters are determined to have converged;
- providing the sum as input to a third adaptive filter to generate a third estimated signal output;
- subtracting the third estimated signal from the detected signal to determine a second output signal;
- and
- adapting third filter coefficients of the third adaptive filter to converge.
10 more elements
Dependencies are inferred from the claim text.