Sound recording device, sound recording program, and sound recording method

Abstract

To obtain a high-quality sound collection result in target direction sound collection processing.SOLUTION: The present invention relates to a sound collection device. The sound collection device comprises: non-target area sound extraction means for acquiring a beamformer output of a microphone array based on an input signal to be supplied from the microphone array including a first microphone and a second microphone, and extracting non-target direction sound from a direction other than a target direction to a target area based on the acquired beamformer output; and target direction sound extraction means for executing selection processing to select one of a component of the first input signal from the first microphone and a component of the second input signal from the second microphone for each frequency, and acquiring a target direction sound extraction signal which is obtained by extracting target direction sound from the target direction by subtracting a component of non-target direction sound extracted by the non-target area sound extraction means from the component selected in the selection processing for each frequency.SELECTED DRAWING: Figure 1

Claims

1. A non-target area sound extraction means that, based on an input signal supplied from a microphone array comprising a first microphone and a second microphone, acquires the beamformer output of the microphone array and extracts non-target direction sounds from directions other than the target direction to the target area based on the acquired beamformer output, A target direction sound extraction means performs a selection process for each frequency to select either a component of the first input signal from the first microphone or a component of the second input signal from the second microphone, and for each frequency, subtracts the component of non-target direction sound extracted by the non-target area sound extraction means from the component selected in the selection process to obtain a target direction sound extraction signal that extracts the target direction sound from the target direction. A sound-collecting device characterized by having the following features.

2. The sound collection device according to claim 1, characterized in that the target direction sound extraction means selects the component with the greater intensity from the component of the first input signal and the component of the second input signal for each frequency in the selection process.

3. Computers, A non-target area sound extraction means that, based on an input signal supplied from a microphone array comprising a first microphone and a second microphone, acquires the beamformer output of the microphone array and extracts non-target direction sounds from directions other than the target direction to the target area based on the acquired beamformer output, A target direction sound extraction means performs a selection process for each frequency to select either a component of the first input signal from the first microphone or a component of the second input signal from the second microphone, and for each frequency, subtracts the component of non-target direction sound extracted by the non-target area sound extraction means from the component selected in the selection process to obtain a target direction sound extraction signal that extracts the target direction sound from the target direction. A sound acquisition program characterized by its ability to function in this way.

4. In the sound collection method performed by a sound collection device, The sound-collecting device comprises means for extracting sounds from non-target areas and means for extracting sounds from target directions. The non-target area sound extraction means acquires the beamformer output of the microphone array based on the input signal supplied from the microphone array comprising a first microphone and a second microphone, and extracts non-target direction sounds from directions other than the target direction to the target area based on the acquired beamformer output. The target direction sound extraction means performs a selection process for each frequency to select either a component of the first input signal from the first microphone or a component of the second input signal from the second microphone. For each frequency, it subtracts the component of non-target direction sound extracted by the non-target area sound extraction means from the component selected in the selection process to obtain a target direction sound extraction signal that extracts the target direction sound from the target direction. A sound collection method characterized by the following.

5. For each of a plurality of microphone arrays, each having two microphones, a non-target area sound extraction means acquires the beamformer output of the microphone array and extracts non-target direction sounds from directions other than the target direction to the target area based on the acquired beamformer output, For each microphone array, a selection process is performed to select either a component of the first input signal from the first microphone or a component of the second input signal from the second microphone for each frequency, and for each frequency, a target direction sound extraction means is performed to subtract the component of non-target direction sound extracted by the non-target area sound extraction means from the component selected in the selection process to obtain a target direction sound extraction signal that extracts the target direction sound from the target direction, Area sound collection means for collecting sound from the target area using the target direction sound extraction signal of each of the microphone arrays, and A sound-collecting device characterized by having the following features.

6. Computers, For each of a plurality of microphone arrays, each having two microphones, a non-target area sound extraction means acquires the beamformer output of the microphone array and extracts non-target direction sounds from directions other than the target direction to the target area based on the acquired beamformer output, For each microphone array, a selection process is performed to select either a component of the first input signal from the first microphone or a component of the second input signal from the second microphone for each frequency, and for each frequency, a target direction sound extraction means is performed to subtract the component of non-target direction sound extracted by the non-target area sound extraction means from the component selected in the selection process to obtain a target direction sound extraction signal that extracts the target direction sound from the target direction, Area sound collection means for collecting sound from the target area using the target direction sound extraction signal of each of the microphone arrays, and A sound acquisition program characterized by its ability to function in this way.

7. In the sound collection method performed by a sound collection device, The sound collection device comprises a means for extracting non-target area sounds, a means for extracting target direction sounds, and an area sound collection means. The non-target area sound extraction means acquires the beamformer output of each of the multiple microphone arrays, each of which has two microphones, and extracts non-target direction sounds from directions other than the target direction to the target area based on the acquired beamformer output. The target direction sound extraction means performs a selection process for each microphone array, for each frequency, to select either a component of the first input signal from the first microphone or a component of the second input signal from the second microphone, and for each frequency, subtracts the component of non-target direction sound extracted by the non-target area sound extraction means from the component selected in the selection process to obtain a target direction sound extraction signal that extracts the target direction sound from the target direction. The area sound collection means collects sound from the target area, with the target area as the sound source, using the target direction sound extraction signal from each of the microphone arrays. A sound collection method characterized by the following.

Images