Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for reducing audio noise are disclosed. In one aspect, a method includes the actions of receiving first audio data of a user utterance. The actions further include determining an energy level of second audio data being outputted by the loudspeaker. The actions further include selecting a model from among (i) a first model that is trained using first audio data samples that each encode speech from one speaker and (ii) a second model that is trained using second audio data samples that each encode speech from either one speaker or two speakers. The actions further include providing the first audio data as an input to the selected model. The actions further include receiving processed first audio data. The actions further include outputting the processed first audio data.

Abstract: 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.

Abstract: 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.

Abstract: A method (400) includes receiving a microphone audio signal (132) and a playout audio signal (112), and determining a frequency representation (324) of the microphone audio signal and a frequency representation of the playout audio signal. For each frequency representation, the method also includes determining features (302) based on the frequency representation. Each feature corresponds to a pair of frequencies (342) of the frequency representation and a period of time between the pair of frequencies. The method also includes determining that a match (212) 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 (222) between the first feature and the second feature corresponds to an echo within the microphone audio signal.

Abstract: 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.

Abstract: A method (400) includes receiving a microphone audio signal (132) and a playout audio signal (112), and determining a frequency representation (324) of the microphone audio signal and a frequency representation of the playout audio signal. For each frequency representation, the method also includes determining features (302) based on the frequency representation. Each feature corresponds to a pair of frequencies (342) of the frequency representation and a period of time between the pair of frequencies. The method also includes determining that a match (212) 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 (222) between the first feature and the second feature corresponds to an echo within the microphone audio signal.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for reducing audio noise are disclosed. In one aspect, a method includes the actions of receiving first audio data of a user utterance. The actions further include determining an energy level of second audio data being outputted by the loudspeaker. The actions further include selecting a model from among (i) a first model that is trained using first audio data samples that each encode speech from one speaker and (ii) a second model that is trained using second audio data samples that each encode speech from either one speaker or two speakers. The actions further include providing the first audio data as an input to the selected model. The actions further include receiving processed first audio data. The actions further include outputting the processed first audio data.