Back to results
US 9,319,783 B1Granted
Attenuation of output audio based on residual echo
AssigneeAmazon Technologies, Inc.
InventorsWilliam Folwell Barton, Amit Chhetri
Life of this patent
Estimating the term…
Abstract
Residual echo that remains after an echo cancellation process may interfere with speech recognition. If near-end speech is detected in an audio input signal, a controller may attenuate the audio playback signal. Too much attenuation may disturb playback, whereas too little attenuation may not improve speech recognition. Accordingly, features are disclosed for attenuating an audio playback signal based at least in part on residual echo level.
Claims24
A system comprising:
- a speaker configured to produce sound based on an audio playback signal;
- a microphone configured to detect the produced sound as an audio input signal;
- one or more processors in communication with the speaker and the microphone, the one or more processors configured to:
- perform acoustic echo cancellation on the audio input signal using a set of acoustic echo cancellation coefficients to generate an echo-reduced signal, the echo-reduced signal comprising an initial residual echo;
- perform residual echo suppression on the echo-reduced signal using a set of residual echo suppression parameters to generate a residual-echo-reduced signal, the residual-echo-reduced signal comprising a reduced residual echo;
- determine that near-end speech is present in the audio input signal;
- determine an attenuation amount for the audio playback signal based at least in part on at least one of the initial residual echo, the acoustic echo cancellation coefficients, the reduced residual echo, or the residual echo suppression parameters;
- and
- attenuate the audio playback signal based on the attenuation amount, wherein the one or more processors are further configured to attenuate the audio playback signal by:
- determining spectral information corresponding to the audio input signal;
- determining, based at least in part on the spectral information, a first portion of the audio playback signal associated with speech and a second portion of the audio playback signal not associated with speech;
- decreasing an amplitude of the first portion of the audio playback signal;
- and
- increasing an amplitude of the second portion of the audio playback signal.
11 more elements
A method comprising:
- receiving an audio playback signal;
- receiving an audio input signal;
- performing echo cancellation on the audio input signal using the audio playback signal and a set of acoustic echo cancellation parameters to generate an echo-reduced signal, the echo-reduced signal comprising a residual echo;
- determining that near-end speech of a user is present in the audio input signal;
- determining an attenuation amount for the audio playback signal;
- and
- attenuating the audio playback signal based on the attenuation amount, wherein attenuating the audio playback signal comprises:
- determining spectral information corresponding to at least a portion of the audio input signal;
- determining, based at least in part on the spectral information, a portion of the audio playback signal associated with speech;
- and
- decreasing an amplitude of the portion of the audio playback signal associated with speech.
8 more elements
One or more non-transitory computer-readable storage media comprising computer-executable instructions to:
- receive an audio playback signal;
- receive an audio input signal;
- process the audio input signal using a set of processing parameters to generate a processed signal;
- determine that near-end speech is present in the audio input signal;
- determine an attenuation amount for the audio playback signal;
- and
- attenuate the audio playback signal based on the attenuation amount, wherein attenuating the audio playback signal comprises:
- determining, based at least in part on spectral information corresponding to at least a portion of the audio input signal, a portion of the audio playback signal associated with speech;
- and
- modifying an amplitude of the portion of the audio playback signal associated with speech.
7 more elements
Dependencies are inferred from the claim text.