Sound collection control device, sound collection control method, and sound collection control program

Abstract

To control directivity of a directional microphone on the basis of a position of a registered person.SOLUTION: A sound collecting control apparatus 100 includes: a determination part 133 for determining whether a characteristic of an acquired voice signal matches a characteristic of a voice signal of a registrant; a specification part 134 for specifying an arrival direction of the acquired voice signal when a determination result is true; and a directivity control part 135 for controlling directivity of a directional microphone 10 on the bass of the specified arrival direction.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a sound collection control device, a sound collection control method, and a sound collection control program, and in particular to a technique for controlling the directivity of a directional microphone. [Background technology]

[0002] A technology has been proposed that extracts a voice signal of a specific person from an audio signal of a television etc. and adjusts the volume of the extracted audio signal. Patent Document 1 discloses a technology that registers voiceprint data of performers in a broadcast program and adjusts the volume of the audio signal of a desired performer.

[0003] On the other hand, Patent Document 2 discloses a technology for estimating the direction from which a specific sound is coming and collecting the sound so as to emphasize the sound coming from the estimated direction. Here, the specific sound is a keyword, a whistle, a clap, etc. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2013-121096 A [Patent Document 2] JP 2020-18015 A Summary of the Invention [Problem to be solved by the invention]

[0005] When a person working from home participates in a remote conference, there is a risk that the voices of people other than the conference participants may be picked up by the microphone. However, with the technology described in Patent Document 1, the picked up voices must be separated and processed into the voices of the participants and the voices of the other people, which causes a problem of a large processing load. Note that Patent Document 2 cannot solve this problem unless the participants make specific sounds such as clapping. Therefore, we provide a sound collection control device, a sound collection control method, and a sound collection control program that can control the directivity of the microphone based on the direction in which a registered person is located. [Means for solving the problem]

[0006] The present invention provides a sound collection control device that includes a judgment unit that judges whether the characteristics of an acquired voice signal match the characteristics of a registered user's voice signal, an identification unit that identifies the direction of arrival of the acquired voice signal when the judgment result is true, and a directionality control unit that controls the directionality of a directional microphone based on the identified direction of arrival.

[0007] The present invention provides a sound collection control method including the steps of: determining whether or not characteristics of an acquired voice signal match characteristics of a registered user's voice signal; if the determination result is true, identifying the direction of arrival of the acquired voice signal; and controlling the directivity of a directional microphone based on the direction of arrival.

[0008] The present invention provides a sound collection control program that causes a computer to execute the steps of determining whether the characteristics of an acquired voice signal match the characteristics of a registered user's voice signal, and if the determination result is true, identifying the direction from which the acquired voice signal comes, and controlling the directionality of a directional microphone based on the direction of arrival. Effect of the Invention

[0009] According to the present invention, it is possible to provide a sound collection control device, a sound collection control method, and a sound collection control program capable of controlling the directivity of a directional microphone based on the direction in which a registered person is located. [Brief description of the drawings]

[0010] [Figure 1] 1 is a block diagram showing a configuration of a sound collection control device according to an embodiment; [Diagram 2] 4 is a flowchart showing a flow of a sound collection control method according to the embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] First, the configuration of a sound collection control device 100 according to an embodiment will be described with reference to Fig. 1. The sound collection control device 100 is, for example, an information terminal such as a PC (Personal Computer), a mobile device, or a tablet terminal. Note that in the embodiment, the sound collection control device 100 is mainly described as being used in a remote conference, but the sound collection control device 100 does not have to be used in a remote conference. The sound collection control device 100 may be a device that collects a person's voice, such as a telephone, chat, or singing device.

[0012] The sound collection control device 100 is connected to a directional microphone 10. The directional microphone 10 is a microphone whose directivity can be controlled. The directional microphone 10 may be configured with a plurality of microphones, such as a stereo microphone or a microphone array. In such a case, the directivity of the directional microphone 10 is controlled by adjusting the volume of the audio signal collected by each microphone. The directional microphone 10 may also be configured with a single microphone having directionality. In such a case, the directivity of the directional microphone 10 is controlled by mechanically rotating the single microphone. The directional microphone 10 may be included in the sound collection control device 100.

[0013] The sound collection control device 100 is also connected to other terminals (not shown) via a network N. The network N may be wired or wireless. The other terminals (not shown) are information terminals operated by other participants of the remote conference. The sound collection control device 100 may be connected to a server (not shown) that provides services related to the remote conference.

[0014] The sound collection control device 100 includes a storage unit 110, a communication unit 120, and a control unit 130. The storage unit 110 is a storage device or storage medium such as a hard disk, a flash memory, or an SD card. The storage unit 110 may also include a volatile storage device such as a RAM (Random Access Memory) that is a storage area for temporarily holding information. The storage unit 110 may be provided in part or in whole in another terminal connected to the network N, and may be connected to the control unit 130 via the communication unit 120.

[0015] The storage unit 110 stores feature data 111. The feature data 111 is data on features of the voice signal of a registered person (registered person). A registered person is, for example, a participant of a remote conference. The feature data 111 may be extracted using a technology such as voiceprint analysis. The feature data 111 is, for example, high and low frequency, loudness and softness of sound volume, or changes over time thereof, but is not limited to this. The feature data 111 of a plurality of people may be stored in the storage unit 110. The storage unit 110 may also store a discriminator (classifier) ​​for discriminating a speaker from the features of a voice signal.

[0016] The storage unit 110 also stores a program (not shown) that implements a control method according to the embodiment, and an application 112 for conducting a remote conference.

[0017] The communication unit 120 is a communication interface with the network N. The communication unit 120 receives an audio signal from another device or a server. The received audio signal is output to a speaker (not shown) provided in the sound collection control device 100. The communication unit 120 also transmits an audio signal acquired by an audio acquisition unit 131 (described later) to another device or the like via the network N.

[0018] The control unit 130 controls the directivity of the directional microphone 10. The functions of the control unit 130 may be realized by loading a program (not shown) into a memory such as a RAM and executing the program by a processor (not shown). The control unit 130 includes a voice acquisition unit 131, a voice analysis unit 132, a determination unit 133, an identification unit 134, a directivity control unit 135, and a remote conference control unit 136.

[0019] The voice acquiring unit 131 acquires voice collected by the directional microphone 10. More specifically, the voice acquiring unit 131 acquires a voice signal by collecting surrounding voice with a microphone and converting the voice signal into an electrical signal. The voice acquiring unit 131 may be configured to include the directional microphone 10. The voice acquiring unit 131 may acquire a voice signal collected by a microphone (not shown) that does not have directionality.

[0020] The voice analysis unit 132 analyzes the features of the voice signal acquired by the voice acquisition unit 131 using a technique such as voiceprint analysis. The voice analysis unit 132 extracts, for example, high and low frequencies, strong and weak sound volumes, or changes in these over time as feature data. The voice analysis unit 132 may extract feature data of the voice signal using other known techniques.

[0021] The determination unit 133 determines whether the feature data extracted by the voice analysis unit 132 matches the feature data 111 stored in the storage unit 110. That is, the determination unit 133 determines whether the voice signal acquired by the voice acquisition unit 131 is a voice uttered by a registrant. The determination unit 133 is also called a speaker identification unit. The determination unit 133 may compare the feature data 111 with the features of the extracted voice signal to perform voice authentication.

[0022] Furthermore, the determination unit 133 may perform the determination using a discriminator (classifier) ​​constructed by supervised learning using the features of the voice signal of the registrant and the features of the voice signal of other people. This enables the determination unit 133 to accurately determine whether the voice signal acquired by the voice acquisition unit 131 is a voice signal of the registrant. The discriminator is also called a machine learning dictionary. The storage unit 110 stores a discriminator constructed by machine learning. When the determination unit 133 performs the determination using the discriminator, the voice analysis unit 132 is not necessarily an essential component and may be omitted.

[0023] When the determination result by the determination unit 133 is true, the identification unit 134 identifies the direction of arrival of the voice signal acquired by the voice acquisition unit 131. When the determination result is true, it may be the case that the voice authentication is successful. The identification unit 134 is also called a speaker position identification unit.

[0024] For example, the determination unit 134 may seek the directivity of the directional microphone 10 and determine the direction in which the loudness (volume) of the audio signal picked up by the directional microphone 10 is maximum. In addition, when the directional microphone 10 is composed of multiple microphones, the determination unit 134 may compare the delay amount of each audio signal picked up by each microphone and acquired by the audio acquisition unit 131, and determine that the direction in which the microphone that picked up the audio signal with the smallest delay is located is the direction from which the audio signal arrives. The determination unit 134 may determine the arrival direction of the picked-up audio signal by using other known techniques. The direction of the audio signal can also be said to be the direction in which the speaker of the acquired audio signal is located.

[0025] The directivity control unit 135 controls the directivity of the directional microphone 10 based on the direction of arrival of the voice signal identified by the identification unit 134. For example, it is assumed that the characteristic data 111 of the participants of the remote conference are registered in advance in the storage unit 110. As a result, the sound collection control device 100 controls the directivity of the directional microphone 10 to face the direction of the remote conference participant's voice identified by the identification unit 134 during the remote conference, thereby making it possible to collect the voice signal of the remote conference participant and to suppress the collection of voices other than those registered, such as voices of children, sounds of daily life, and external noises such as wind noise.

[0026] Conversely, the feature data 111 of the person not participating in the conference may be registered in advance in the storage unit 110. In such a case, the directivity control unit 135 controls the directivity of the directional microphone 10 so as to face a direction different from the direction specified by the specification unit 134. This enables the sound collection control device 100 to suppress collection of voice signals from people other than the participants of the remote conference, such as children, during the remote conference.

[0027] The remote conference control unit 136 accesses a URL or the like (conference room) where the remote conference is held, and transmits the audio signal acquired by the audio acquisition unit 131 via the communication unit 120. The function of the remote conference control unit 136 is realized by the processor (not shown) executing the application 112.

[0028] The control unit 130 may control the directivity of the directional microphone 10 when a specific application 112 is running, and may not control the directivity of the directional microphone 10 when the specific application 112 is not running. The case where the application 112 is not running refers to, for example, when the remote conference control unit 136 is not operating. This allows the control unit 130 to control the directivity of the directional microphone 10 only when necessary. The specific application 112 is not limited to an application for remote conferences, but may be an application for voice communication with the outside, such as a telephone or chat tool, an application for karaoke or singing, an application for voice recording, or an application for picking up voice.

[0029] Furthermore, the control unit 130 may acquire, as the conference information, participant information, the conference name, or conference room information of a remote conference held using the application 112. In such a case, the control unit 130 controls the directivity of the directional microphone 10 when the conference information satisfies a predetermined condition, and does not control the directivity of the directional microphone 10 when the conference information does not satisfy the predetermined condition. This allows the control unit 130 to efficiently control the directivity of the directional microphone 10 only when a specific person is participating in the remote conference, when the remote conference is held by specific members or for a specific purpose, or when the remote conference is held using a specific conference room (URL, etc.).

[0030] Furthermore, the control unit 130 may obtain information indicating whether or not an audio signal is being transmitted to the outside from the communication unit 120, and may control the directivity of the directional microphone 10 when the audio signal is being transmitted to the outside, and may control not to control the directivity of the directional microphone 10 when the audio signal is not being transmitted to the outside. This allows the sound collection control device 100 to efficiently control the directivity of the directional microphone 10, since the directivity of the directional microphone 10 is controlled toward the voice direction of the registered person only when the audio acquired by the audio acquisition unit 131 is being transmitted to the outside, for example, when communication with the outside such as a remote conference or telephone is being performed.

[0031] 2 is a flowchart illustrating a flow of a sound collection control method according to an embodiment. It is assumed that feature data 111 of voice signals of people participating in a remote conference is registered in the storage unit 110. First, the control unit 130 determines whether or not an application 112 is running in the sound collection control device 100 (step S101).

[0032] The sound collection control device 100 ends the process if the application 112 is not running (No in step S101), and proceeds to the process of step S102 if the application 112 is running (Yes in step S101). In step S101, the sound collection control device 100 may determine whether or not the participant information of the remote conference satisfies a predetermined condition, or may determine whether or not the communication unit 120 is transmitting an audio signal to the outside. Note that the sound collection control method according to the embodiment may also include a case in which the process of step S102 is started without making the determination in step S101.

[0033] Next, the voice acquisition unit 131 of the control unit 130 acquires a voice signal collected by the directional microphone 10 or a microphone having no directionality (step S102). Next, the voice analysis unit 132 of the control unit 130 extracts feature data from the voice signal acquired in step S102 (step S103). Next, the determination unit 133 of the control unit 130 determines whether the feature data extracted in step S103 is feature data 111 of the voice signal of the registrant (step S104). Here, the determination unit 133 may perform authentication using the feature data 111 stored in the storage unit 110, or may perform classification using a classifier constructed by machine learning.

[0034] If the extracted feature is not a feature of the voice signal of a registered person (No in step S104), the control unit 130 proceeds to the end determination process in step S107. On the other hand, if the extracted feature data matches the feature data of the voice signal of a registered person (Yes in step S104), the sound collection control device 100 identifies the arrival direction of the voice signal collected by the directional microphone 10 (step S105).

[0035] Next, the sound collection control device 100 controls the directivity of the directional microphone 10 to face the arrival direction identified in step S105 (step S106). If sound is collected by a microphone having no directionality in step 102, it is preferable to control the directivity of the directional microphone 10 and stop or suppress collection of sound by the microphone having no directionality. Then, the process proceeds to the end determination process in step S107. Note that if the voice of a child or the like is registered in advance, the directivity control unit 135 of the sound collection control device 100 directs the directivity of the directional microphone 10 in a direction different from the direction identified in step S105.

[0036] After step S106 ends, the control unit 130 proceeds to the end determination process of step S107. In step S107, the sound collection control device 100 determines whether or not to end the series of processes. If it is not determined that the series of processes should be ended (No in step S107), the sound collection control device 100 returns to step S102, acquires an audio signal, and continues the process. If it is determined that the series of processes should be ended (Yes in step S107), the sound collection control device 100 ends the process. For example, the sound collection control device 100 may determine that the process should be ended when the application 112 is not running. The sound collection control device 100 may determine that the process should be ended when the communication unit 120 is not transmitting an audio signal to the outside.

[0037] The sound collection control device according to the embodiment controls the directivity of the microphone when the sound signal acquired by the sound acquisition unit 131 is the sound signal of a registered person, and does not control the directivity of the microphone when the sound signal acquired by the sound acquisition unit 131 is not the sound signal of a registered person. Therefore, the sound collection control device according to the embodiment can control the directivity of the directional microphone according to the position of the registered person. Since the directivity of the directional microphone is not controlled when the acquired sound is not the sound of a registered person, there is no risk of controlling the directivity of the microphone in the direction of the sound of another person.

[0038] For example, when participating in a remote conference while working from home, a directional microphone is pointed in the direction of the participants to pick up sound, so that the participants' voices are picked up properly while producing natural sound without the unnaturalness that comes from electrically processing sounds other than those of the participants, and it is possible to suppress the pickup of sounds such as children's voices, household sounds, and external noise that come from directions other than those of the participants.

[0039] The above-mentioned program can be stored and supplied to a computer using various types of non-transitory computer readable media. The non-transitory computer readable media includes various types of tangible storage media. The non-transitory computer readable media includes, for example, a magnetic recording medium, a magneto-optical recording medium, a CD-ROM (Read Only Memory), a CD-R, a CD-R / W, and a semiconductor memory. The magnetic recording medium is, for example, a flexible disk, a magnetic tape, a hard disk drive, etc. The magneto-optical recording medium is, for example, a magneto-optical disk, etc. The semiconductor memory is, for example, a mask ROM, a PROM (Programmable ROM), an EPROM (Erasable PROM), a flash ROM, a RAM (Random Access Memory), etc. The program may also be supplied to a computer by various types of transitory computer readable media. Examples of the transitory computer readable medium include an electric signal, an optical signal, and an electromagnetic wave. The temporary computer-readable medium can supply the program to the computer via a wired communication path such as an electric wire or an optical fiber, or via a wireless communication path.

[0040] The sound collection control device according to the above embodiment may be mounted on a vehicle, and may be incorporated in, for example, a drive recorder or a car navigation device. This makes it possible to point a directional microphone toward a registered person, regardless of which seat the person is sitting in the vehicle. Therefore, even inside a vehicle where the influence of noise is large, it is possible to collect the clear voice of the person. The device may also be mounted on a household electrical appliance or electronic device that can be controlled by voice. [Explanation of symbols]

[0041] 10 Directional Microphone 100 Sound collection and control device 110 Storage section 111 Feature Data 112 Applications 120 Communications Department 130 Control section 131 Voice Acquisition Unit 132 Audio Analysis Section 133 Judgment section 134 Specific part 135 Directivity control section 136 Remote conference control unit N Network

Claims

1. a determination unit that determines whether or not the characteristics of the acquired voice signal match those of a voice signal of a registrant; an identification unit that identifies the arrival direction of the acquired voice signal when the determination result is true; a directivity control unit that controls a directivity of a directional microphone based on the identified arrival direction; a communication unit that transmits the acquired voice signal to an outside; Equipped with If the determination result is not true, the process related to the control of the directivity of the directional microphone is terminated. When the acquired audio signal is being transmitted to the outside by the communication unit, the directivity of the directional microphone is controlled, and when the acquired audio signal is not being transmitted to the outside by the communication unit, the directivity of the directional microphone is not controlled. Sound pickup control device.

2. The sound collection control device further includes: Whether or not a predetermined application is running in the sound collection control device is set as a condition for whether or not to control the directivity of the directional microphone. The sound collection control device according to claim 1 .

3. A computer comprising: A step of determining whether the characteristics of the acquired voice signal match those of the registered person. Tep and If the determination result is true, determining the direction of arrival of the acquired voice signal; controlling a directivity of a directional microphone based on the direction of arrival; a communication step of transmitting the acquired voice signal to an outside; A sound collection control method for executing the above, If the determination result is not true, the process related to the control of the directivity of the directional microphone is terminated. A sound collection control method, characterized in that, when the acquired audio signal is transmitted to the outside by the communication step, the directivity of the directional microphone is controlled, and when the acquired audio signal is not transmitted to the outside by the communication step, the directivity of the directional microphone is not controlled.

4. On the computer, A step of determining whether or not the characteristics of the acquired voice signal match the characteristics of the voice signal of a registered person; If the determination result is true, determining the direction of arrival of the acquired voice signal; controlling a directivity of a directional microphone based on the direction of arrival; a communication step of transmitting the acquired voice signal to an outside; A sound collection control program for executing If the determination result is not true, the process related to the control of the directivity of the directional microphone is terminated. A sound collection control program, characterized in that, when the acquired audio signal is transmitted to the outside by the communication step, a directivity of the directional microphone is controlled, and when the acquired audio signal is not transmitted to the outside by the communication step, the directivity of the directional microphone is not controlled.

Images