Wideband doubletalk detection for optimization of acoustic echo cancellation

Patent metadata

Field	Value
Publication number	US-12598261-B2
Application number	US-202318477039-A
Country	US
Kind code	B2
Filing date	Sep 28, 2023
Priority date	Sep 28, 2022
Publication date	Apr 7, 2026
Grant date	Apr 7, 2026

Abstract

Acoustic echo cancellation systems and methods are provided that improve the quality of the audio transmitted from by an audio device a near end to a far end when a doubletalk condition is present, including allowing certain subbands of an echo-cancelled signal to be less attenuated by overriding certain gains of subbands of the echo-cancelled audio signal in a non-linear processor, and compressing and applying makeup gain to a remote audio signal.

First claim

The invention claimed is:

1 . A device, comprising:
one or more processors configured to:
determine an initial echo-cancelled audio signal based on a microphone signal and a remote audio signal;
determine that a doubletalk condition is present in one or more first subbands of the initial echo-cancelled audio signal;
determine that a doubletalk condition is not present in one or more second subbands of the initial echo-cancelled audio signal;
override one or more gains of the one or more second subbands of the initial echo-cancelled audio signal to generate a final echo-cancelled audio signal; and
transmit the final echo-cancelled audio signal to a remote location.

2 . The device of claim 1 , wherein any of the one or more processors is configured to determine that the doubletalk condition is present by:
determining the one or more gains of the one or more first subbands of the initial echo-cancelled audio signal; determining that a doubletalk condition is present in one or more subbands of the microphone signal and the remote audio signal; and determining that the doubletalk condition is present when the doubletalk condition is present in the one or more subbands of the microphone signal and the remote audio signal.

3 . The device of claim 2 , wherein any of the one or more processors is configured to determine that the doubletalk condition is present by:
determining that the doubletalk condition is present in each subband of the microphone signal and the remote audio signal; based on a number of subbands having the doubletalk condition present satisfying a predetermined doubletalk condition criteria, determining that the doubletalk condition is present; and based on the number of subbands having the doubletalk condition present not satisfying the predetermined doubletalk condition criteria, determining that the doubletalk condition is not present.

4 . The device of claim 2 , wherein any of the one or more processors is configured to determine that the doubletalk condition is present in the one or more subbands of the microphone signal and the remote audio signal based on a ratio of an energy of the microphone signal to an energy of the remote audio signal.

5 . The device of claim 2 , wherein any of the one or more processors is configured to override one or more gains of the one or more second subbands of the initial echo-cancelled audio signal by:
increasing, in a non-linear processor, the one or more gains of the one or more second subbands of the initial echo-cancelled audio signal that do not have presence of the doubletalk condition.

6 . The device of claim 1 , wherein any of the one or more processors is further configured to:
when the doubletalk condition is determined to be present:
compress the remote audio signal to generate a compressed remote audio signal; and
play the compressed remote audio signal on a loudspeaker.

7 . The device of claim 6 , wherein any of the one or more processors is configured to compress the remote audio signal by limiting a volume of one or more subbands of the remote audio signal.

8 . The device of claim 6 , wherein any of the one or more processors is configured to compress the remote audio signal by reducing a dynamic range of the remote audio signal.

9 . The device of claim 6 , wherein any of the one or more processors is further configured to, when the doubletalk condition is determined to be present, apply a makeup gain to the compressed remote audio signal to generate a compressed gain-corrected remote audio signal.

10 . A method, comprising:
determining an initial echo-cancelled audio signal based on a microphone signal and a remote audio signal; determining that a doubletalk condition is present in one or more first subbands of the initial echo-cancelled audio signal; determining that a doubletalk condition is not present in one or more second subbands of the initial echo-cancelled audio signal; overriding one or more gains of the one or more second subbands of the initial echo-cancelled audio signal to generate a final echo-cancelled audio signal; and transmitting the final echo-cancelled audio signal to a remote location.

11 . The method of claim 10 , wherein determining that the doubletalk condition is present comprises:
determining the one or more gains of the one or more first subbands of the initial echo-cancelled audio signal; determining that a doubletalk condition is present in one or more subbands of the microphone signal and the remote audio signal; and determining that the doubletalk condition is present when the doubletalk condition is present in the one or more subbands of the microphone signal and the remote audio signal.

12 . The method of claim 11 , wherein determining that the doubletalk condition is present comprises:
determining that the doubletalk condition is present in each subband of the microphone signal and the remote audio signal; based on a number of subbands having the doubletalk condition present satisfying a predetermined doubletalk condition criteria, determining that the doubletalk condition is present; and based on the number of subbands having the doubletalk condition present not satisfying the predetermined doubletalk condition criteria, determining that the doubletalk condition is not present.

13 . The method of claim 11 , wherein determining that the doubletalk condition is present in the one or more subbands of the microphone signal and the remote audio signal is based on a ratio of an energy of the microphone signal to an energy of the remote audio signal.

14 . The method of claim 11 , wherein overriding the one or more gains of the one or more second subbands of the initial echo-cancelled audio signal comprises:
increasing, in a non-linear processor, the one or more gains of the one or more second subbands of the initial echo-cancelled audio signal that do not have presence of the doubletalk condition.

15 . The method of claim 10 , further comprising:
when the doubletalk condition is determined to be present:
compressing the remote audio signal to generate a compressed remote audio signal; and
playing the compressed remote audio signal on a loudspeaker.

16 . The method of claim 15 , wherein compressing the remote audio signal comprises limiting a volume of one or more subbands of the remote audio signal.

17 . The method of claim 15 , wherein compressing the remote audio signal comprises reducing a dynamic range of the remote audio signal.

18 . The method of claim 15 , further comprising when the doubletalk condition is determined to be present, applying a makeup gain to the compressed remote audio signal to generate a compressed gain-corrected remote audio signal.

19 . A system, comprising:
a microphone configured to generate a microphone signal; a loudspeaker configured to generate sound based on a remote audio signal, wherein the microphone and the loudspeaker are located in a same housing; and one or more processors configured to:
determine an initial echo-cancelled audio signal based on a microphone signal and a remote audio signal;
determine that a doubletalk condition is present in one or more first subbands of the initial echo-cancelled audio signal; and
determine that a doubletalk condition is not present in one or more second subbands of the initial echo-cancelled audio signal;
override one or more gains of the one or more second subbands of the initial echo-cancelled audio signal to generate a final echo-cancelled audio signal; and
compress the r

Assignees

• Shure Acquisition Holdings Inc

Inventors

• Lethem Wesley
• Sconza Justin Joseph
• Rieger Christopher George

Classifications

• H04M9/082 Primary

  using echo cancellers (echo cancellers per se H04B3/23) · CPC title