JP2013183315A - Hands-free speech apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To selectively send voices of a speaking person to a speech partner through simple arithmetic processing.SOLUTION: A hands-free speech apparatus 100 comprises a receiver 110, an ultrasonic speaker 120, a microphone array 130, a BSS (Blind Source Separation) processor 140 and a transmitter 150. The receiver 110 receives a reception voice. The ultrasonic speaker 120 outputs the reception voice received by the receiver 110. The microphone array 130 collects sounds. The BSS processor 140 extracts a transmission voice from the sounds collected by the microphone array 130. The transmitter 150 transmits the transmission voice extracted by the BSS processor 140.

Description

  The present invention relates to a hands-free call device.

  As a hands-free call device capable of making a call even if the device and the speaker are not close to each other, for example, there are technologies of Patent Documents 1 and 2.

  In the hands-free call device of Patent Document 1, the position of a speaker is detected by image recognition, and the detection result is used for directivity control of a microphone and a speaker. The directivity control of the microphone is performed by selectively using a microphone corresponding to the direction of the speaker among a plurality of microphones, or by changing the direction of the microphone by the driving unit.

  In the hands-free call device of Patent Document 2, the position of a speaker is detected by an ultrasonic sensor, and the detection result is used for directivity control of a microphone and a speaker. In the technique of Patent Document 2, the directivity of the microphone is increased in the specified direction by controlling the phase and amplitude of the sound input of a plurality of microphones.

  In Patent Document 3, a moving body (a robot or the like) that recognizes the position of a speaker by image recognition, outputs a sound in a specific direction by a parametric speaker array, and collects sound by a microphone having directivity toward the front. ).

  Further, in Patent Document 4, when an incoming call is received at a telephone terminal, a sound signal is obtained by collecting sound from a plurality of microphones provided in the telephone terminal, and a signal obtained by removing the ringing tone component from the sound signal is estimated as a noise signal. The technology is described.

JP 2010-232755 A JP 2001-359187 A International Publication No. 2005/076661 Pamphlet JP 2009-153053 A

  In a hands-free call device, it is necessary to use many microphones in order to selectively transmit a speaker's voice to a call partner. However, since a mounting space is limited in a mobile communication terminal device or the like, it may be difficult to mount many microphones. In Patent Document 1, when the direction of the microphone is changed by the drive unit, the drive unit is required, so that the mounting space is increased.

  In addition, if the arithmetic processing in the hands-free call device is complicated, there is an adverse effect such as an increase in power consumption. Therefore, it is desirable that the arithmetic processing is as simple as possible.

An object of the present invention is to provide a hands-free call device capable of solving any of the following problems.
The first problem is to selectively transmit the voice of the speaker to the other party with a simple arithmetic process.
The second problem is to selectively transmit the speaker's voice to the other party using as few microphones as possible.

The present invention includes a receiver that receives a received voice;
An ultrasonic speaker for outputting the received voice received by the receiver;
A microphone array that collects audio,
A BSS (Blind Source Separation) processing unit for extracting a transmitted voice from the voice collected by the microphone array;
A transmission unit for transmitting the transmission voice extracted by the BSS processing unit;
There is provided a hands-free communication device characterized by comprising:

In addition, the present invention provides a receiver that receives a received voice;
An ultrasonic speaker for outputting the received voice received by the receiver;
An AMNOR (Adaptive Microphone-array for NOISE Reduction) microphone array that collects sound;
A transmission unit that generates a transmission voice based on the voice collected by the microphone array, and transmits the transmission voice;
There is provided a hands-free communication device characterized by comprising:

  According to the present invention, the voice of the speaker can be selectively transmitted to the other party with a simple arithmetic process. Alternatively, according to the present invention, the voice of the speaker can be selectively transmitted to the other party using as few microphones as possible.

It is a block diagram which shows the structure of the hands-free call device which concerns on 1st Embodiment. It is a block diagram which shows the more concrete structure of the hands-free call apparatus which concerns on 1st Embodiment. It is a typical top view which shows the structure of an ultrasonic speaker. It is typical sectional drawing which shows the structure of each ultrasonic transducer | vibrator of an ultrasonic speaker. It is a schematic diagram for demonstrating operation | movement of 1st Embodiment. It is a figure which shows the characteristic of the resonant frequency of the ultrasonic transducer | vibrator of an ultrasonic speaker. It is a schematic diagram for demonstrating the example of directivity control of an ultrasonic speaker. It is a flowchart for demonstrating operation | movement of the hands-free call apparatus in case a speaker is plural. It is a block diagram which shows the structure of the hands-free call apparatus which concerns on 2nd Embodiment. It is a block diagram which shows the more concrete structure of the hands-free call apparatus which concerns on 2nd Embodiment. It is a schematic diagram for demonstrating operation | movement of 2nd Embodiment.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In all the drawings, the same components are denoted by the same reference numerals, and description thereof will be omitted as appropriate.

[First Embodiment]
FIG. 1 is a block diagram showing a configuration of a hands-free communication device 100 according to the first embodiment.

  The hands-free communication device 100 according to the present embodiment includes a receiver 110, an ultrasonic speaker 120, a microphone array 130, a BSS (Blind Source Separation) processor 140, and a transmitter 150. The receiver 110 receives the received voice. The ultrasonic speaker 120 outputs the reception voice received by the reception unit 110. The microphone array 130 collects sound. The BSS processing unit 140 extracts the transmitted voice from the voice collected by the microphone array 130. The transmitter 150 transmits the transmitted voice extracted by the BSS processor 140.

  The BSS processing unit 140 performs a process called blind signal source separation (blind sound source separation). This process is a process of separating audio data derived from each sound source from the audio data collected in an environment where a plurality of sound sources are mixed.

  If there is any noise source other than the voice of the speaker who is the user of the hands-free communication device 100, the microphone array 130 collects noise from the noise source in addition to the voice of the speaker. The BSS processing unit 140 separates the voice (voice data) collected by the microphone array 130 into a speaker's voice and noise. Further, the BSS processing unit 140 extracts the speaker's voice as the transmitted voice from the separated voice. The transmitting unit 150 transmits the voice of the speaker extracted by the BSS processing unit 140 to the calling device of the other party as the transmitted voice.

  Thus, according to the hands-free call device 100 according to the present embodiment, the BSS processing unit 140 extracts the transmitted voice from the voice collected by the microphone array 130. Therefore, the voice of the speaker can be selectively transmitted to the other party.

Here, in the hands-free communication device 100, the received voice is output by the ultrasonic speaker 120. For this reason, the received voice is demodulated at a certain distance from the ultrasonic speaker 120 (and hence from the hands-free call device 100), and the received voice is reproduced as an audible sound. Therefore, it can be suppressed that the microphone array 130 picks up the acoustic echo of the received voice. As a result, the BSS processing unit 140 does not need to perform either the process of separating the acoustic echo of the received voice or the process of determining the acoustic echo as noise. That is, the processing of the BSS processing unit 140 is simplified.
In short, the voice of the speaker can be selectively transmitted to the other party with a simple arithmetic process.

  Hereinafter, the first embodiment will be described in more detail.

  FIG. 2 is a block diagram showing a more specific configuration of the hands-free communication device 100 according to the first embodiment.

  The hands-free call device 100 is a device for making a call with another call device by wireless communication or wired communication. The receiving unit 110 receives a received voice transmitted from another call device. The transmitter 150 transmits the transmitted voice to another communication device.

  The hands-free call device 100 is preferably a mobile phone or other mobile communication terminal device, but may be a landline phone or the like.

  As shown in FIG. 2, the hands-free call device 100 includes an input unit 20, an imaging unit 170, a position specifying unit 180, and an image determination unit 190 in addition to the above configuration.

  The input unit 20 oscillates (outputs) ultrasonic waves from the ultrasonic speaker 120 by inputting a modulation signal for a parametric speaker to the ultrasonic speaker 120. The ultrasonic wave output from the ultrasonic speaker 120 is demodulated at a position somewhat away from the hands-free call device 100 and reproduced as an audible sound. As a result, the speaker is in a hands-free state, that is, in a state where the speaker's mouth is not brought close to the hands-free call device 100 (typically, the hands-free call device 100 is not held by hand). The sound transmitted from the device (not shown) can be heard.

  The input unit 20 includes a directivity control unit 21 that controls the directivity of the ultrasonic speaker 120.

  The imaging unit 170 is a camera that captures an image of a speaker or the like.

  The position specifying unit 180 specifies the position of the speaker based on the imaging result obtained by the imaging unit 170. That is, the position specifying unit 180 performs predetermined image recognition processing such as face recognition on the image captured by the image capturing unit 170, so that the position (direction and distance) of the speaker based on the hands-free call device 100 is determined. Determine.

  The directivity control unit 21 of the input unit 20 controls the directivity of the ultrasonic speaker 120 so that the sound is reproduced at the position of the speaker specified by the position specifying unit 180.

  The image determination unit 190 determines the speaker's utterance timing based on the imaging result obtained by the imaging unit 170. More specifically, the image determination unit 190 determines the movement of the mouth for the image (that is, the speaker's image) at the position specified by the position specifying unit 180. The timing at which the mouth performs a predetermined operation (such as when opening and closing is repeated) is determined to be the speaker's utterance timing.

  The BSS processing unit 140 extracts the transmitted voice from the voice collected by the microphone array 130 at the utterance timing of the speaker determined by the image determination unit 190. For this reason, the transmitted voice, that is, the voice of the speaker can be extracted efficiently.

  The microphone array 130 includes a plurality of microphones 131 (see FIG. 5).

  FIG. 3 is a schematic plan view showing the configuration of the ultrasonic speaker 120.

  As shown in FIG. 3, the ultrasonic speaker 120 includes a plurality of ultrasonic transducers 30 that output ultrasonic waves. That is, the ultrasonic speaker 120 includes a transducer array including a plurality of ultrasonic transducers 30. The input unit 20 causes each ultrasonic transducer 30 to emit ultrasonic waves by inputting a modulation signal to each ultrasonic transducer 30.

  The plurality of ultrasonic transducers 30 are arranged in a matrix, for example. In the example of FIG. 3, a total of 16 ultrasonic transducers 30 are arranged in a matrix in 4 columns and 4 columns. However, the number and arrangement of the ultrasonic transducers 30 included in the ultrasonic speaker 120 are not limited to this example.

  The ultrasonic speaker 120 includes, for example, a frame-shaped support unit 70 that supports the plurality of ultrasonic transducers 30.

  FIG. 4 is a schematic cross-sectional view showing the configuration of each ultrasonic transducer 30 of the ultrasonic speaker 120.

The ultrasonic transducer 30 includes, for example, a sheet-like vibrating member 62 and a transducer 63. The vibrator 63 is a piezoelectric vibrator and is attached to one surface of the vibration member 62. The entire surface of the vibrator 63 facing the vibration member 62 is fixed to the vibration member 62 with an adhesive. As a result, the entire surface of one surface of the vibrator 63 is restrained by the vibration member 62.
The edge of the vibration member 62 is supported by the support portion 70.

  The material constituting the vibration member 62 is not particularly limited as long as it is a material having a high elastic modulus with respect to the vibrator 63 which is a brittle material such as metal or resin, but phosphor bronze, stainless steel, etc. from the viewpoint of workability and cost. Is preferred. The vibration member 62 vibrates due to vibration generated from the vibrator 63, and oscillates a sound wave having a frequency of 20 kHz or more, for example. The vibrator 63 also oscillates, for example, a sound wave having a frequency of 20 kHz or more by vibrating itself.

  The vibrator 63 includes a piezoelectric ceramic (not shown) and electrode films respectively formed on both surfaces of the piezoelectric ceramic.

  The input unit 20 functions as an oscillator that oscillates the transducer 63 of each ultrasonic transducer 30. The input unit 20 generates an electric signal input to the vibrator 63, that is, a modulation signal (oscillation signal) for a parametric speaker. The transport wave of the modulation signal is, for example, an ultrasonic wave having a frequency of 20 kHz or higher, and specifically, an ultrasonic wave having a frequency of 100 kHz, for example. The input unit 20 controls the ultrasonic transducer 30 so that a predetermined oscillation output is obtained.

  A parametric speaker emits ultrasonic waves (transport waves) subjected to AM modulation, DSB modulation, SSB modulation, and FM modulation from each of a plurality of oscillation sources into the air, and due to nonlinear characteristics when the ultrasonic waves propagate into the air. , To make an audible sound appear. Non-linear here means that the flow changes from laminar flow to turbulent flow when the Reynolds number indicated by the ratio between the inertial action and the viscous action of the flow increases. Since the sound wave is slightly disturbed in the fluid, it propagates nonlinearly. Particularly in the ultrasonic frequency band, the nonlinearity of the sound wave can be easily observed. And when an ultrasonic wave is radiated in the air, harmonics accompanying the nonlinearity of the sound wave are remarkably generated. The sound wave is a dense state where the density of the molecular density is generated in the air. When it takes time for air molecules to recover from compression, air that cannot be recovered after compression collides with air molecules that continuously propagate, and a shock wave is generated. An audible sound is generated by this shock wave. The audible sound, that is, the received voice, is reproduced at a position away from the ultrasonic speaker 120 (and thus from the hands-free communication device 100) by a certain distance.

  For example, the input unit 20 can control the plurality of ultrasonic transducers 30 such that the plurality of ultrasonic transducers 30 oscillate at individual phases. Directivity can be enhanced or directivity can be arbitrarily controlled by causing the plurality of ultrasonic transducers 30 to oscillate at individual phases. However, the input unit 20 may oscillate a plurality of ultrasonic transducers 30 with the same phase.

  FIG. 5 is a schematic diagram for explaining the operation of the first embodiment.

  For example, as shown in FIG. 5, it is assumed that there are two speakers, a first speaker 210 and a second speaker 220, as speakers. Further, it is assumed that another speaker 50 exists outside the hands-free call device 100 and noise is emitted from the speaker 50.

  First, the imaging unit 170 captures an image including the first speaker 210 and the second speaker 220. The position specifying unit 180 specifies the positions of the first speaker 210 and the second speaker 220 based on the imaging result obtained by the imaging unit 170. The directivity control unit 21 of the input unit 20 controls the directivity of the ultrasonic speaker 120 so that the received voice is reproduced at the positions of the first speaker 210 and the second speaker 220 specified by the position specifying unit 180. To do. That is, the directivity of the ultrasonic speaker 120 is controlled so that the received voice is reproduced only in the vicinity of the first speaker 210 and the second speaker 220. As a result, it is possible to listen to the received voice in the audible area 40 shown in FIG.

  The microphone array 130 collects sound (voice data) including the utterances of the first speaker 210 and the second speaker 220 and noise from the speaker 50.

  The BSS processing unit 140 separates sound originating from each sound source (first speaker 210, second speaker 220, and speaker 50) from the sound collected by the microphone array 130.

  Further, the BSS processing unit 140 determines noise from a plurality of voices separated from each other, and extracts a voice excluding the noise as a transmission voice. Therefore, the transmitter 150 can selectively transmit the voices of the first speaker 210 and the second speaker 220 to the telephone device of the other party.

  Here, the BSS processing unit 140 is configured to recognize, for example, a sound obtained by removing a ringtone from a sound collected by the microphone array 130 at the time of an incoming call as noise. Thereby, when the sound is emitted from the speaker 50 before the incoming call, the sound can be easily recognized as noise.

  Further, the image determination unit 190 determines the utterance timing of the first speaker 210 and the second speaker 220 based on the imaging result by the imaging unit 170. Then, the BSS processing unit 140 extracts the transmitted voice from the voice collected at the utterance timing determined by the image determination unit 190. For this reason, transmission voice can be extracted efficiently.

  Next, an example of directivity control of the ultrasonic speaker 120 will be described with reference to FIGS. 6 and 7. FIG. 6 is a diagram showing the characteristic of the resonance frequency of the ultrasonic transducer 30 of the ultrasonic speaker 120. FIGS. 7A, 7 </ b> B, and 7 </ b> C are schematic diagrams for explaining an example of directivity control of the ultrasonic speaker 120.

  In the case of the present embodiment, the directivity control unit 21 changes the resonance frequency of the ultrasonic speaker 120 by adjusting a modulation signal input from the input unit 20 to the ultrasonic speaker 120, for example. The directivity of the speaker 120 is controlled. That is, the directivity of the ultrasonic speaker 120 is controlled by changing the resonance frequency of each ultrasonic transducer 30 by adjusting the modulation signal input from the input unit 20 to each ultrasonic transducer 30.

  As shown in FIG. 6, the ultrasonic transducer 30 has a plurality of resonance frequencies. That is, the ultrasonic transducer 30 includes a carrier frequency a1 (for example, 20 kHz), a harmonic frequency a2 that is twice the carrier frequency a1 (for example, 40 kHz), and a third harmonic frequency that is three times the carrier frequency a1. a3 (for example, 60 kHz).

  The directivity of the ultrasonic speaker 120 becomes higher as the resonance frequency becomes higher, that is, as the frequency of the carrier wave becomes higher. In addition, the directivity of the ultrasonic speaker 120 becomes lower as the resonance frequency and the frequency of the carrier wave become lower. FIG. 7A schematically shows the directivity (angle range of the region where the audible sound is demodulated) when the resonance frequency of the ultrasonic transducer 30 is the carrier frequency a1. Similarly, FIG. 7B schematically shows the directivity when the resonance frequency of the ultrasonic transducer 30 is the harmonic frequency a2, and FIG. 7C shows the resonance frequency of the ultrasonic transducer 30 3. The directivity when the harmonic frequency is a3 is schematically shown. 7A to 7C, θ1> θ2> θ3.

  Therefore, by appropriately selecting the resonance frequency from the carrier frequency a1, the harmonic frequency a2, and the third harmonic frequency a3 according to the positional relationship between the hands-free communication device 100 and the speaker and the number of speakers, The directivity of the sound wave speaker 120 can be appropriately controlled.

  Next, the processing of the BSS processing unit 140 when there are a plurality of speakers as described above will be described in more detail.

  As described above, the BSS processing unit 140 separates the sound originating from each sound source from the sound collected by the microphone array 130. For this reason, the BSS processing unit 140 also separates the voice of the first speaker 210 and the voice of the second speaker 220 from each other. For this reason, the voice of the first speaker 210 and the voice of the second speaker 220 are synthesized (synchronized and added) to generate a transmission voice.

  FIG. 8 is a flowchart for explaining the operation of the hands-free call device 100 when there are a plurality of speakers.

  First, sound is collected by the microphone array 130 (step S11). Next, the BSS processing unit 140 separates the sound originating from each sound source (step S12). Next, the BSS processing unit 140 determines whether there are a plurality of speakers (step S13). This process is performed by determining whether or not there are a plurality of voices among a plurality of voices separated from each other. Furthermore, as an aid to this processing, a determination result as to whether or not there are a plurality of persons in the image captured by the imaging unit 170 may be used.

  If there are a plurality of speakers (Y in step S13), a voice to be transmitted is synthesized by synthesizing the voices (voices) of the plurality of speakers (step S14). On the other hand, when there is only one speaker, the voice (voice) of the speaker is set as the transmitted voice (step S15).

According to the first embodiment as described above, the hands-free call device 100 includes the receiving unit 110 and the ultrasonic speaker 120 that outputs the received voice received by the receiving unit 110. Furthermore, the hands-free call device 100 includes a microphone array 130 that collects voice, a BSS processing unit 140 that extracts transmitted voice from the voice collected by the microphone array 130, and a transmission that is extracted by the BSS processing unit 140. And a transmitter 150 for transmitting speech.
Therefore, the voice of the speaker can be selectively transmitted to the other party.
Here, in the hands-free communication device 100, the received voice is output by the ultrasonic speaker 120. For this reason, the received voice is demodulated at a certain distance from the ultrasonic speaker 120 (and hence from the hands-free call device 100), and the received voice is reproduced as an audible sound. Therefore, it can be suppressed that the microphone array 130 picks up the acoustic echo of the received voice. As a result, the BSS processing unit 140 does not need to perform either the process of separating the acoustic echo of the received voice or the process of determining the acoustic echo as noise. That is, the processing of the BSS processing unit 140 is simplified.
In short, the voice of the speaker can be selectively transmitted to the other party with a simple arithmetic process.

  The hands-free call device 100 includes an imaging unit 170 that captures an image, and an image determination unit 190 that determines a speaker's utterance timing based on an imaging result obtained by the imaging unit 170. Then, the BSS processing unit 140 extracts the transmitted voice from the voice collected at the utterance timing determined by the image determination unit 190. Therefore, the transmission voice extraction process can be performed efficiently.

  The BSS processing unit 140 extracts utterances by a plurality of speakers as a part of each transmitted speech, and generates the transmitted speech by synthesizing the extracted utterances by the plurality of speakers. Therefore, when there are a plurality of speakers, the voices of the plurality of speakers can be selectively transmitted to the other party.

  The hands-free call device 100 includes an imaging unit 170, a position specifying unit 180 that specifies the position of a speaker based on an imaging result obtained by the imaging unit 170, and a directivity control unit 21. The directivity control unit 21 controls the directivity of the ultrasonic speaker so that sound is reproduced at the position of the speaker specified by the position specifying unit 180. Therefore, the voice can be selectively reproduced at the position of the speaker.

[Second Embodiment]
FIG. 9 is a block diagram showing a configuration of the hands-free call device 200 according to the second embodiment. FIG. 10 is a block diagram showing a more specific configuration of the hands-free call device 200 according to the second embodiment. FIG. 11 is a schematic diagram for explaining the operation of the second embodiment.

  The hands-free communication device 100 according to the present embodiment includes a receiver 110, an ultrasonic speaker 120, a microphone array 230, and a transmitter 150. The receiver 110 receives the received voice. The ultrasonic speaker 120 outputs the reception voice received by the reception unit 110. The microphone array 230 collects sound. The microphone array 230 is of an AMPOR (Adaptive Microphone-array for Noise Reduction) system. The transmitter 150 generates a transmission voice based on the voice collected by the microphone array 230 and transmits the transmission voice.

  The AMNOR type microphone array 230 reduces the directivity of the microphone array 230 with respect to the direction of noise. A directivity valley is formed in the noise direction. As a result, the voice of the speaker can be selectively transmitted from the transmitter 150 to the other party. The AMNOR microphone array 230 has high sensitivity in the target sound direction. However, the AMNOR microphone array 230 does not form a sharp directivity in the target sound direction.

  Here, in the AMNOR microphone array 230, the more noise sources, the more microphones 131 (see FIG. 11) are required. This is because the number of directivity valleys that can be formed is (the number of microphones 131 minus 1).

On the other hand, in the hands-free call device 200 according to the present embodiment, the received voice is output by the ultrasonic speaker 120. For this reason, the received voice is demodulated at a position away from the ultrasonic speaker 120 (and thus from the hands-free call device 200) by a certain distance, and the received voice is reproduced as an audible sound. Therefore, it can be suppressed that the microphone array 230 picks up the acoustic echo of the received voice. As a result, the microphone array 230 need not perform either the process of determining the acoustic echo as noise or the process of reducing the directivity with respect to the direction of the acoustic echo. That is, the processing of the microphone array 230 is simplified.
Further, since the number of noise sound sources is reduced as compared with the case of using a normal speaker that outputs an audible sound as the speaker, the number of microphones 131 required by the microphone array 230 can be reduced.
In short, it is possible to selectively transmit the voice of the speaker to the other party with a simple arithmetic process using as few microphones as possible.

  Hereinafter, a more detailed description of the second embodiment will be given focusing on differences from the first embodiment.

  Similar to the hands-free call device 100, the hands-free call device 200 is a device for making a call with another call device by wireless communication or wired communication. The receiving unit 110 receives a received voice transmitted from another call device. The transmitter 150 transmits the transmitted voice to another communication device. The hands-free call device 200 is preferably a mobile phone or other mobile communication terminal device, but may be a landline phone or the like.

  As shown in FIG. 10, the hands-free call device 200 according to the present embodiment does not have the image determination unit 190 and the BSS processing unit 140 in the first embodiment.

  The microphone array 230 includes a plurality of microphones 131 (FIG. 11) and an AMNOR processing unit 235 (FIG. 10). The AMNOR processing unit 235 performs a process of reducing the directivity of the microphone array 230 with respect to the direction of the noise source (hereinafter referred to as noise source). That is, directivity valleys 241 and 242 (FIG. 11) are formed in the direction of the noise source.

  Hereinafter, the operation of the present embodiment will be described with reference to FIG.

  In AMNOR, it is necessary (precondition) that the target sound direction is known and that a signal containing only noise can be received.

  Also in this embodiment, the imaging unit 170 captures an image including the first speaker 210 and the second speaker 220. The position specifying unit 180 specifies the positions of the first speaker 210 and the second speaker 220 as the target sound direction based on the imaging result by the imaging unit 170. The AMNOR processing unit 235 does not decrease the directivity of the microphone array 230 for the target sound direction, and decreases the directivity of the microphone array 230 for directions other than the target sound direction.

  The microphone array 230 determines that a voice from a direction different from the directions of the first speaker 210 and the second speaker 220 specified by the position specifying unit 180 is noise. For example, when there is a person who is not captured in the field of view of the imaging unit 170 (a speaker 250 other than the speaker in FIG. 11), the AMNOR processing unit 235 determines that the person (the sound source in that direction) is a noise source. be able to.

  Note that the AMNOR processing unit 235 determines only persons (for example, the first speaker 210 and the second speaker 220) for which facial image information is registered in advance as speakers, and other persons (other than the speakers). The speaker 250) may be determined as a noise source.

  In addition, the AMNOR processing unit 235 is configured to recognize, as noise, a sound obtained by removing the ringtone from the sound collected by the microphone array 230 at the time of the incoming call. Thereby, when the sound is emitted from the speaker 50 before the incoming call, the AMNOR processing unit 235 can determine the speaker 50 as a noise source.

  Thereby, the AMNOR processing unit 235 controls the microphone array 230 so that directivity valleys 241 and 242 are formed in the direction of the speaker 50 and the direction of the speaker 250 other than the speaker, respectively. As a result, the sound including the utterances of the first speaker 210 and the second speaker 220 is transmitted from the transmitter 150 to the other party's communication device as the transmitted sound.

According to the second embodiment as described above, the hands-free call device 200 includes the receiving unit 110 and the ultrasonic speaker 120 that outputs the received voice received by the receiving unit 110. The hands-free call device 200 further generates an AMNOR-type microphone array 230 that collects voices, a transmission voice based on the voices collected by the microphone array 230, and a transmission that transmits the transmission voices. And a talk unit 150.
The AMNOR type microphone array 230 reduces the directivity of the microphone array 230 with respect to the direction of noise. As a result, the voice of the speaker can be selectively transmitted from the transmitter 150 to the other party.
In addition, since the received voice is output from the ultrasonic speaker 120, the received voice is demodulated at a certain distance from the ultrasonic speaker 120 (and hence from the hands-free call device 100), and the received voice is reproduced as an audible sound. The Therefore, it can be suppressed that the microphone array 230 picks up the acoustic echo of the received voice. As a result, the microphone array 230 need not perform either the process of determining the acoustic echo as noise or the process of reducing the directivity with respect to the direction of the acoustic echo. That is, the processing of the AMNOR processing unit 235 is simplified.
Further, since the number of noise sound sources is reduced as compared with the case of using a normal speaker that outputs an audible sound as the speaker, the number of microphones 131 required by the microphone array 230 can be reduced. Therefore, since the microphone array 230 can be easily mounted in a limited mounting space, the hands-free call device 200 according to the present embodiment can be easily realized as a mobile communication terminal device.

20 Input unit 21 Directivity control unit 30 Ultrasonic transducer 40 Audible region 50 Speaker 62 Vibrating member 63 Vibrator 70 Support unit 100 Hands-free communication device 110 Reception unit 120 Ultrasonic speaker 130 Microphone array 131 Microphone 140 BSS processing unit 150 Talking section 170 Imaging section 180 Position specifying section 190 Image determining section 200 Hands-free call device 210 First speaker (speaker)
220 Second speaker (speaker)
230 Microphone array 235 AMNOR processing unit 241 Directional valley 242 Directional valley 250 Speaker other than speaker a1 Carrier frequency a2 Harmonic frequency a3 Third harmonic frequency

Claims (7)

A receiver for receiving the received voice;
An ultrasonic speaker for outputting the received voice received by the receiver;
A microphone array that collects audio,
A BSS (Blind Source Separation) processing unit for extracting a transmitted voice from the voice collected by the microphone array;
A transmission unit for transmitting the transmission voice extracted by the BSS processing unit;
A hands-free communication device comprising: An imaging unit that captures an image;
An image determination unit for determining a speaker's utterance timing based on a result of imaging by the imaging unit;
Have
The hands-free call device according to claim 1, wherein the BSS processing unit extracts the transmitted voice from a voice collected at the utterance timing determined by the image determination unit.   The BSS processing unit extracts utterances by a plurality of speakers as a part of the transmitted speech, and generates the transmitted speech by synthesizing the extracted utterances by the plurality of speakers. The hands-free call device according to claim 1 or 2. A receiver for receiving the received voice;
An ultrasonic speaker for outputting the received voice received by the receiver;
An AMNOR (Adaptive Microphone-array for NOISE Reduction) microphone array that collects sound;
A transmission unit that generates a transmission voice based on the voice collected by the microphone array, and transmits the transmission voice;
A hands-free communication device comprising: An imaging unit that captures an image;
A position specifying unit for specifying the position of the speaker based on the imaging result of the imaging unit;
Have
The said microphone array makes the directivity with respect to the direction different from the direction of the said speaker specified by the said position specific | specification part lower than the directivity of the direction of the said speaker. Hands-free call device. An imaging unit that captures an image;
A position specifying unit for specifying the position of the speaker based on the imaging result of the imaging unit;
A directivity control unit that controls directivity of the ultrasonic speaker so that sound is reproduced at the position of the speaker specified by the position specifying unit;
The hands-free call device according to claim 1, wherein the hands-free call device is provided.   The directivity control unit controls the directivity of the ultrasonic speaker by changing a resonance frequency of the ultrasonic speaker by adjusting a modulation signal input to the ultrasonic speaker. The hands-free call device according to claim 6.

Images