Google Inc.

Ivo Creusen

Daniel Johansson

Alessio Bazzica

Marcus Wirebrand

Mans Gustaf Sebastian Ullberg

Alexandre Loiko

Samuel Martin Zackrisson

Per Åhgren

Tina Le Grand

A method includes receiving a microphone audio signal and a playout audio signal, and determining a frequency representation of the microphone audio signal and a frequency representation of the playout audio signal. For each frequency representation, the method also includes determining features based on the frequency representation. Each feature corresponds to a pair of frequencies of the frequency representation and a period of time between the pair of frequencies. The method also includes determining that a match occurs between a first feature based on the frequency representation of the microphone audio signal and a second feature based on the frequency representation of the playout audio signal, and determining that a delay value between the first feature and the second feature corresponds to an echo within the microphone audio signal.

Echo detection

A method () includes receiving a microphone audio signal () and a playout audio signal (), and determining a frequency representation () of the microphone audio signal and a frequency representation of the playout audio signal. For each frequency representation, the method also includes determining features () based on the frequency representation. Each feature corresponds to a pair of frequencies () of the frequency representation and a period of time between the pair of frequencies. The method also includes determining that a match () occurs between a first feature based on the frequency representation of the microphone audio signal and a second feature based on the frequency representation of the playout audio signal, and determining that a delay value () between the first feature and the second feature corresponds to an echo within the microphone audio signal.

Methods and systems detect residual echo in an audio signal in a non-causal manner. A non-causal residual echo detector allows accurate detection of any residual echo from an AEC (acoustic echo canceller). The non-causally detected residual echo may be utilized to perform various actions such as selecting between different causal AECs, tuning a causal AEC, and/or characterizing the performance of the causal AEC. Aspects leverage an understanding that the information about residual echo being present in a signal does not have to be causal in the sense that it is acceptable and even advantageous to know some time afterwards (e.g., after a signal has undergone acoustic echo cancellation and is output for further processing or transmission) that a residual echo was present in the signal. Such an approach is different from the case of an AEC, which needs to know right away, with a fairly low signal delay involved, whether residual echo is present.

Residual echo detection based on non-causality

Growing community of inventors

Stockholm, Canada

Ivo Creusen