Beamforming method, beamforming system

Abstract

To provide a method, device, and system for achieving appropriate beamforming.SOLUTION: A beamforming method of performing beamforming on input signals corresponding to sounds input to a plurality of microphones in a binaural beamformer 1 of a binaural listening device 100 receiving inputs through two channels using the minimum variance distortionless response (MVDR) includes: a filter update step of calculating a coefficient based on a result of passing the input signals to MVDR filters 24-1 to 24-4 provided with predetermined designs, and switching FIR filters 27-1 to 27-4 using the coefficient; and a convolution step of convolving the input signals by the FIR filters.SELECTED DRAWING: Figure 4

Claims

1. A beamforming method that performs beamforming using MVDR on input signals corresponding to sound input to multiple microphones, wherein the processor is: A filter update process is performed which involves calculating coefficients based on the result of passing the input signal through an MVDR filter designed to maintain the degree of interaural cross-correlation of noise components contained in the input signal, and switching the FIR filter using the coefficients, A convolution step in which the input signal is convolved with the FIR filter, An adjustment process that allows the degree to be changed. A beamforming method that includes [a specific feature / method].

2. In the beamforming method described in claim 1, The aforementioned MVDR filter is The cost function is expressed by an equation that includes a parameter that controls the degree of the aforementioned degree. In the adjustment process, A beamforming method characterized by allowing the value of the aforementioned parameter to be changed.

3. In the beamforming method according to claim 1 or 2, In the aforementioned convolution process, The input signal is convolved on the first signal path. In the aforementioned filter update process, A beamforming method characterized by calculating the coefficient based on the result of multiplying the frequency domain gain, which is obtained by performing a predetermined calculation for each frequency band on a frequency domain signal corresponding to the input signal on a second signal path branched from the first signal path, by the gain of the MVDR.

4. A beamforming system that uses MVDR to beamform input signals corresponding to sounds input to multiple microphones, A filter update unit calculates coefficients based on the result of passing the input signal through an MVDR filter, which was designed taking into account the degree to which the interaural cross-correlation of the noise components contained in the input signal is maintained, and switches the FIR filter using the coefficients. A convolution section that convolves the input signal using the FIR filter, An adjustment unit that allows the degree to be changed and A beamforming system equipped with [a specific feature / feature].

5. In the beamforming system according to claim 4, The aforementioned MVDR filter is The cost function is expressed by an equation that includes a parameter that controls the degree of the aforementioned degree. The adjustment unit is, A beamforming system characterized by allowing the value of the aforementioned parameter to be changed.

6. In the beamforming system according to claim 4 or 5, The aforementioned folding section is The input signal is convolved on the first signal path. The filter update unit, A beamforming system characterized by calculating the coefficient based on the result of multiplying the frequency domain gain, which performs a predetermined calculation for each frequency band on a frequency domain signal corresponding to the input signal, by the gain of the MVDR on a second signal path branched from the first signal path.

Images