JP2019067423A - Purchase settlement method - Google Patents

Abstract

To provide a purchase settlement method capable of preventing the purchase of merchandise by voice to a speaker who does not want to settle the purchase of merchandise.SOLUTION: A purchase settlement method for purchasing an article by voice acquires voice information, analyzes the acquired voice information, and acquires a voice command indicating a control instruction to the equipment, specifies the speaker information about the speaker who utters the acquired voice information based on the acquired voice information when the voice command is a voice command relating to the purchase settlement, refers to a table in which the purchasable speaker information and the information required for the purchase settlement are associated with each other, judges whether or not the specified speaker information is the speaker information by which the article can be purchased and the purchase should be settled, and performs the purchase settlement processing by using the voice command and the information required for the purchase settlement when it is determined that the specified speaker information is the speaker information by which the article can be purchased and the purchase should be settled.SELECTED DRAWING: Figure 19

Description

  The present disclosure relates to a device control method for controlling a plurality of devices by voice, a display control method for controlling a display device for displaying information on a plurality of devices controlled based on voice from a speaker, and purchasing an article by voice Related to the purchase settlement method.

  BACKGROUND Conventionally, there are voice recognition systems that operate and control various devices in a home by voice input. In such a speech recognition system, in addition to the control of the device by voice input, a service that interacts with the device in a pseudo manner by making a voice response to the user's voice instruction has been proposed (for example, , Patent Document 1).

Unexamined-Japanese-Patent No. 2002-182679

  However, in the above speech recognition system, all devices can be controlled similarly regardless of which speaker speaks. Therefore, even a speaker who does not want to control the device can give voice instructions to the device. As described above, in the conventional voice recognition system, it has not been considered to prevent the speaker not wanting to control the device to control the device by voice. Also, conventionally, no speech recognition system has been considered which can purchase an article safely by speech recognition.

  One aspect of the present disclosure provides a purchase settlement method that can prevent the purchase of a product from being settled by voice for a speaker who does not want to settle the purchase of the product.

  A purchase settlement method according to one aspect of the present disclosure is a purchase settlement method for purchasing an article by voice, acquiring speech information, analyzing the acquired speech information, and acquiring a speech command indicating a control instruction to a device If the voice command is a voice command related to purchase settlement, the speaker information on the speaker who utters the acquired voice information is specified based on the acquired voice information, and the speaker information capable of purchase settlement (B) referring to a table associated with information necessary for purchase settlement to determine whether the specified speaker information is speaker information available for purchase settlement, and the specified speaker information indicates purchase settlement When it is determined that the information is possible speaker information, purchase settlement processing is performed using the voice command and information necessary for the purchase settlement.

  Note that these general or specific aspects may be realized by a system, method, integrated circuit, computer program, or recording medium, and any of the system, apparatus, method, integrated circuit, computer program, and recording medium It may be realized by any combination.

  According to the present disclosure, when the speaker information on the speaker who does not want to settle the purchase of the product is specified, the purchase of the product is not made to be settled by voice to the speaker who does not want to settle the purchase of the product. be able to.

FIG. 1 is a diagram showing an overall configuration of a speech recognition system in the present embodiment. It is a figure which shows the example in which an apparatus maker corresponds to a data center operating company. It is a figure which shows the example which an equipment maker and / or a management company correspond to a data center operating company. It is a figure which shows the whole structure of the speech recognition system in the modification of this Embodiment. It is a figure which shows the structure of the apparatus in Embodiment 1 of this indication. It is a figure which shows the structure of the control apparatus in Embodiment 1 of this indication. It is a figure which shows the structure of the database in Embodiment 1 of this indication. It is a figure which shows the function structure of the speech recognition system in Embodiment 1 of this indication. It is a flowchart which shows operation | movement of the speech recognition system in Embodiment 1 of this indication. It is a sequence diagram for demonstrating the operation | movement with respect to the 1st apparatus of the speech recognition system in Embodiment 1 of this indication. It is a sequence diagram for demonstrating the operation | movement with respect to the 2nd apparatus of the speech recognition system in Embodiment 1 of this indication. It is a figure which shows the structure of the control processing part in Embodiment 1 of this indication, and a database. It is a figure which shows an example of the table which shows the relationship between the audio | voice information currently recorded on the database, and the age estimated. It is a figure which shows an example of the table which shows the relationship between the audio | voice model memorize | stored in the database, and a control object apparatus. It is a figure which shows an example of the table which shows the relationship between the audio | voice command currently recorded on the database, and a control object apparatus. It is a figure which shows an example of the table which matched the audio | voice command and control content which were memorize | stored in the memory of an apparatus. It is a figure which shows an example of the display screen which displays the apparatus which can be operated by audio | voice by a list. It is a figure for demonstrating the operation | movement which sets the speaker operated by an audio | voice for every apparatus. It is a figure which shows an example of the display screen displayed when the speaker operated by an audio | voice is set for every apparatus. It is a figure which shows the function structure of the speech recognition system in Embodiment 2 of this indication. It is a flowchart which shows operation | movement of the speech recognition system in Embodiment 2 of this indication. It is a sequence diagram for demonstrating the purchase payment process of the speech recognition system in Embodiment 2 of this indication. It is a figure which shows the structure of the control apparatus in Embodiment 2 of this indication, and a server. It is a figure which shows an example of the 1st table which shows the relationship between the audio | voice information currently recorded on the database, estimated age, and estimated sex. It is a figure which shows an example of the table which shows the relationship between the user number memorize | stored in the purchase payment database, audio | voice information, age, and credit card information. It is a figure which shows the whole image of the service which the speech recognition system in service type 1 (company data center type | mold cloud service) provides. It is a figure which shows the overall image of the service which the speech recognition system in service type 2 (IaaS type cloud service) provides. It is a figure which shows the overall image of the service which the speech recognition system in service type 3 (PaaS utilization type | mold cloud service) provides. It is a figure which shows the overview of the service which the speech recognition system in service type 4 (SaaS utilization type | mold cloud service) provides.

(Findings that formed the basis of this disclosure)
Patent Document 1 describes a voice recognition system in which a plurality of devices connected to a network exchange information to respond to and operate an appropriate device in response to a voice command from a user.

  However, in the speech recognition system of Patent Document 1, the same response and operation are uniformly performed regardless of who issues the speech command. That is, even if a third person who is not a person in the home who owns a plurality of devices to be controlled speaks, each device can be operated like a person in the home. Also, for example, it is conceivable that a child operates a cooking device that handles fire by a voice command or operates a device that does not need to be operated.

  In addition, in a voice recognition system that interacts with a plurality of devices as in Patent Document 1, it is expected to construct a voice recognition system which can purchase an article by voice command without being limited to control of devices. . For example, in a scene in which the user interacts with a refrigerator, when the user designates a food having no stock in the refrigerator, the user is asked whether or not to purchase the designated food, and the user selects the food when purchasing. It is conceivable to place an order via a network. Even in such a case, it is not preferable that the voice recognition system is operated by anyone's voice.

  That is, in the conventional voice recognition system such as Patent Document 1, anyone can operate all the devices by voice, and the speaker who does not want to operate the devices is prevented from operating the devices by voice. It was not considered in regard to things. Also, conventionally, no speech recognition system has been considered which can purchase an article safely by speech recognition.

  In order to solve the above problems, a device control method according to an aspect of the present disclosure is a device control method for controlling a plurality of devices by voice, which obtains voice information, analyzes the obtained voice information, and a device. The voice command indicating the control instruction for the voice is acquired, the speaker information on the speaker who utters the acquired voice information is specified based on the obtained voice information, and the plurality of devices are specified based on the voice command and the speaker information. Identify the control target device of

  According to this configuration, the control target device among the plurality of devices is specified based on the voice command and the speaker information, so when speaker information regarding a speaker who does not want to control the device is specified, For a speaker who does not want to control the device, the device can not be controlled by voice.

  Further, in the above aspect, the speaker information preferably includes information on the age of the speaker.

  According to this configuration, it is possible to prevent the device from being controlled by voice according to the age of the speaker. For example, a speaker of a child or an elderly person can be prevented from controlling the device by voice.

  Further, in the above aspect, reference is made to each of a first table that associates and stores speaker information and a candidate device indicating a candidate of a control target device, and a second table that associates and stores a voice command and a candidate device. Preferably, candidate devices that match both the first table and the second table are identified as control target devices.

  According to this configuration, the first table in which the speaker information and the candidate device indicating the candidate of the control target device are associated and stored, and the second table in which the voice command and the candidate device are associated and stored are referred to. Candidate devices that match both the first table and the second table are identified as controlled devices.

  Therefore, among the candidate devices corresponding to the specified speaker information and the candidate devices corresponding to the acquired voice command, the matching candidate device is determined as the control target device, so the control target device can be easily determined. can do.

  Furthermore, in the above aspect, when there is no candidate device that matches both of the first table and the second table, it is preferable to notify that there is no control target device.

  According to this configuration, since it is notified that there is no device to be controlled, the user can know that there is no device to be controlled with respect to the voice command uttered.

  Further, in the above aspect, the third table which associates and stores the speaker information and the use history of the device of the speaker corresponding to the speaker information is referred to, and the use history of the device of the speaker corresponding to the speaker information It is preferable to specify the control target device based on the and the voice command.

  According to this configuration, when there are a plurality of candidate devices, it is possible to determine the device used at the same time in the past as the control target device based on the usage history, and to be used most frequently in the past Can be determined as the control target device.

  Further, in the above aspect, the speaker position information indicating the position of the speaker and the device position information indicating the positions of the plurality of devices are acquired, and the voice command, the speaker information, the speaker position information and the device position information It is preferable to identify a control target device based on

  According to this configuration, the speaker position information indicating the position of the speaker and the device position information indicating the positions of the plurality of devices are acquired, and the voice command, the speaker information, the speaker position information and the device position information are obtained. Since the control target device is specified based on the above, when there are a plurality of candidate devices, the device closest to the position of the speaker can be determined as the control target device.

  Further, in the above aspect, the display device displaying the information on the device is controlled to display the list of the first information on the plurality of devices controllable by the voice, and the control of the device by the voice regarding the specific speaker is performed. It is preferable to control the display device to display the second information indicating the rejection, in the vicinity of the first information corresponding to the device that does not receive the control.

  According to this configuration, a list of first information related to a plurality of controllable audio devices is displayed. Then, second information indicating that the control of the device by voice is not received for the specific speaker is displayed in the vicinity of the first information corresponding to the device not receiving control.

  Therefore, the user can easily confirm the plurality of devices that can be controlled by voice and can easily check the devices that can not be controlled by a specific speaker among the plurality of devices that can be controlled by voice. Can.

  A display control method according to another aspect of the present disclosure is a display control method for controlling a display device that displays information related to a plurality of devices controlled based on a voice from a speaker, Control the display device to display a list of the first information on the device, and correspond to the device not receiving the control, the second information indicating that the control of the device by voice is not received for the specific speaker The display device is controlled to be displayed in the vicinity of the first information.

  According to this configuration, a list of first information related to a plurality of controllable audio devices is displayed. Then, second information indicating that the control of the device by voice is not received for the specific speaker is displayed in the vicinity of the first information corresponding to the device not receiving control.

  Therefore, the user can easily confirm the plurality of devices that can be controlled by voice and can easily check the devices that can not be controlled by a specific speaker among the plurality of devices that can be controlled by voice. Can.

  In the aspect described above, the display device is a touch panel display, and the first information and the second information arranged at a position different from the first information are displayed on the display screen of the display device. The display area displaying the second information is touched by the user, the second information is dragged toward the first information, and the second information is dropped on the first information. And controlling the display device to display the second information in the vicinity of the first information, and when the second information is displayed in the vicinity of the first information, the voice of the device corresponding to the first information It is preferable to change the setting for a specific speaker so as not to accept control by the speaker.

  According to this configuration, the first information and the second information arranged at a position different from the first information are displayed on the display screen of the display device. The display area displaying the second information is touched by the user, the second information is dragged toward the first information, and the second information is dropped on the first information. Information is displayed in the vicinity of the first information. Then, when the second information is displayed in the vicinity of the first information, the setting for the specific speaker is changed so as not to receive the control by the voice of the device corresponding to the first information.

  Therefore, among a plurality of devices that can be controlled by voice, it is possible to easily set a device that does not receive control by a specific speaker.

  A purchase settlement method according to another aspect of the present disclosure is a purchase settlement method for purchasing an article by voice, acquiring speech information, analyzing the acquired speech information, and acquiring a speech command indicating a control instruction to a device If the voice command is a voice command for purchase settlement, the speaker information on the speaker who utters the acquired voice information is specified based on the acquired voice information, and the purchase settlement possible speaker information and the purchase settlement Refer to the table in which the necessary information is associated, and it is determined whether or not the specified speaker information is speaker information that can be purchased and settled, and the speaker information that is identified can be purchased and settled speaker information If it is determined that the purchase settlement process is performed using the voice command and the information necessary for purchase settlement.

  According to this configuration, the table in which the purchase settlement possible speaker information is associated with the information necessary for the purchase settlement is referred to, and it is determined whether the identified speaker information is the purchase settlement possible speaker information. If it is determined that the determined speaker information is the speaker information that can be purchased and settled, the purchase settlement process is performed using the voice command and the information necessary for the purchase and settlement.

  Therefore, when the speaker information on the speaker who does not want to settle the purchase of the product is specified, it is possible not to make the purchase of the product be settled by voice to the speaker who does not want to settle the purchase of the product.

(Overall configuration of speech recognition system)
FIG. 1A is a diagram showing an overall configuration of a speech recognition system according to the present embodiment. FIG. 1B is a diagram showing an example in which the device maker corresponds to a data center operating company. FIG. 1C is a diagram showing an example in which both or one of the device maker and the management company corresponds to the data center operating company.

  The group 100 is, for example, a company, a group, a family, or the like, regardless of the size. The group 100 includes a plurality of devices 101 including a device A, a device B, and a device C, a home gateway 102, a sound collection device 103, and a speaker 104.

  The plurality of devices 101 can be connected to the Internet (for example, a smartphone, a personal computer (PC) or a television), and devices that can not be connected to the Internet by themselves (for example, a lighting device, a washing machine, or a refrigerator) including. The plurality of devices 101 may include devices that can be connected to the Internet via the home gateway 102 even if the devices can not connect to the Internet by themselves. Also, the user 10 uses a plurality of devices 101 in the group 100. The device connectable to the Internet may be connected directly to the control device 111 described later without passing through the home gateway 102. The plurality of devices 101 receive control commands from the control device 111 described later directly or via the home gateway 102.

  The home gateway 102 is connected to some or all of the plurality of devices 101, the sound collection device 103, or the speaker 104, and transmits and receives various information. The home gateway 102 and each device may be connected by wire using a cable or the like, or may be connected wirelessly using Wi-Fi or Bluetooth (registered trademark). In addition, the home gateway 102 is connected to the control device 111 to transmit and receive information.

  The sound collection device 103 acquires the sound in the room of the group 100. The acquired voice is output to the control device 111 via the home gateway 102. Alternatively, the acquired voice is directly output from the sound collection device 103 to the control device 111. The sound collection device 103 is configured by a general microphone, and the installation location of the sound collection device 103 is not particularly limited. It is preferable that the sound collection device 103 be installed on a ceiling, a lighting fixture, or the like that can easily acquire the sound of a room. In addition, the sound collection device 103 may have a function of directing directivity in the direction in which the user 10 is present.

  Moreover, the sound collection apparatus 103 does not need to be fixed in the room, and may be provided in a movable portable terminal such as a smartphone.

  The speaker 104 receives a control command from the control device 111 via the home gateway 102, and outputs a sound according to the received control command. Alternatively, the speaker 104 receives a control command directly from the control device 111, and outputs a sound according to the received control command.

  The home gateway 102 and the speaker 104 are not essential components of the voice recognition system.

  The data center operating company 110 comprises a control device 111. The control device 111 is a virtualization server that cooperates with various devices via the Internet. The control device 111 mainly manages huge data (big data) and the like which are difficult to handle with ordinary database management tools and the like. The data center operating company 110 manages data, manages the control device 111, and operates a data center that performs these operations.

  Here, the data center operating company 110 is not limited to a company that only manages data or operates the control device 111. For example, as shown in FIG. 1B, when a device maker developing and manufacturing one of the plurality of devices 101 manages data or manages the control device 111, the device maker It corresponds to the data center operating company 110. Also, the data center operating company 110 is not limited to one company. For example, as shown in FIG. 1C, when a device maker and another management company jointly or share management of data management or control apparatus 111, both or any one of them is the data center operating company 110. Applicable

  The control device 111 is connected to the home gateway 102 via the Internet, and transmits / receives various information to / from a plurality of devices 101 connected to the home gateway 102, the sound collection device 103, the speaker 104, and the like. Alternatively, the control device 111 directly connects to the plurality of devices 101, the sound collection device 103, the speaker 104, and the like via the Internet, and transmits and receives various information. Further, the control device 111 is connected to a server 121 described later to transmit and receive information.

  The service provider 120 comprises a server 121 for providing services to the user 10. The server 121 referred to here includes, for example, a memory in a personal PC, etc., regardless of its size. In addition, the service provider 120 may not include the server 121. The service provider 120 and the data center operating company 110 may be the same company. The server 121 receives information from the control device 111, and performs control and processing related to the service provided to the user. The service may be provided to the users 10 in the group 100 or may be provided to the users 20 outside the group 100.

  Although the control device 111 is disposed outside the group 100 in FIG. 1A, the present disclosure is not particularly limited to this, and as shown in FIG. It may be arranged.

  FIG. 2 is a diagram showing an overall configuration of a speech recognition system according to a modification of the present embodiment.

  As shown in FIG. 2, the controller 111 may exist in the group 100.

  The above is the description of the overall configuration of the speech recognition system. Each embodiment will be described below. The embodiments described below all show one specific example of the present disclosure. The numerical values, shapes, components, steps, and order of steps and the like described in the following embodiments are an example of the present disclosure, and are not intended to limit the present disclosure. Further, among the components in the following embodiments, components that are not described in the independent claim indicating the highest concept of the present disclosure are described as arbitrary components. Moreover, in all the embodiments, the contents of the respective embodiments can be combined.

Embodiment 1
(Configuration of each device)
FIG. 3 is a diagram showing the configuration of the device in the first embodiment of the present disclosure. FIG. 4 is a diagram showing a configuration of a control device in the first embodiment of the present disclosure. FIG. 5 is a diagram showing the configuration of a database in the first embodiment of the present disclosure.

  First, the configuration of a plurality of devices 101 (device A, device B, device C,...) In FIG. 1A will be described using FIG. 3. Here, the plurality of devices 101 include all devices in the home. The device 101 a is one of the plurality of devices 101. The device 101a includes, for example, various devices such as a television, a BD (Blu-ray (registered trademark) Disc) recorder, an air conditioner, a refrigerator, a washing machine, a microwave, and various cookers (gas stove and IH cooking heater). Not limited The device 101a may be an electric appliance used in the home, including not only home appliances but also home appliances, sensors for measuring and detecting a home environment, and electric vehicles.

  The device 101a includes, for example, a communication unit 201, a control processing unit 202, a memory 203, a microphone 204, a speaker 205, and a display unit 206. However, since the device 101a includes various devices as described above, the configuration of the device 101a is not limited to this. The device 101a may not include, for example, any or all of the memory 203, the microphone 204, the speaker 205, and the display unit 206. In addition, the device 101a may include another configuration not illustrated.

  The communication unit 201 transmits and receives information to and from various devices via a network or via various cables. The communication unit 201 is configured by hardware such as various communication ICs (Integrated Circuits).

  The control processing unit 202 processes and calculates various data (information) received by the communication unit 201, and controls the operation of each component in the device 101. The control processing unit 202 is configured by hardware such as a central processing unit (CPU), for example.

  The memory 203 stores data (information) received by the communication unit 201, data (information) calculated by the control processing unit 202, or a control program. The memory 203 is configured by hardware such as a ROM (Read Only Memory), for example.

  The microphone 204 picks up (acquires) a voice command of a person, for example. The microphone 204 is not particularly limited as long as it is generally mounted on a device. With regard to the speaker 205 and the display unit 206 as well, those generally mounted on a device may be used, and the description will be omitted.

  Next, the configuration of the control device 111 will be described using FIG. The control device 111 includes a communication unit 211, a control processing unit 212, and a database 213. However, the configuration of the control device 111 is not limited to this. Some configurations of the controller 111 may be omitted, or other configurations may be included.

  The communication unit 211 transmits / receives information to / from various devices via a network or via various cables and the like. The communication unit 211 is configured by hardware such as various communication ICs (Integrated Circuits).

  The control processing unit 212 processes and calculates various data (information) received by the communication unit 211, and controls the operation of each component in the control device 111 or in the plurality of devices 101. The control processing unit 212 is configured by hardware such as a central processing unit (CPU), for example.

  The database 213 stores various data. The database 213 may include, for example, the voice of a person acquired by the microphone 204 or the sound collection device 103 of the device 101a, or the use history of the device 101a. The data accumulated in the database 213 is used by the control processing unit 212 for processing of estimating the age or gender of the speaker. Details of the database 213 will be described later. Further, as shown in FIG. 5, the database 213 may associate and accumulate data acquired for each group (family, company or group, etc.). That is, the database 213 may include a database storing data of the first group, a database storing data of the second group, and a database storing data of the third group.

(Functional configuration of speech recognition system)
FIG. 6 is a diagram showing a functional configuration of the speech recognition system in the first embodiment of the present disclosure. As shown in FIG. 6, the speech recognition system includes a speech acquisition unit 301, a speech analysis unit 302, a speaker identification unit 303, a control target device identification unit 304, and a device control unit 305.

  The voice acquisition unit 301 acquires voice information indicating the voice of the speaker that has been emitted around the device. The sound acquisition unit 301 is realized by the sound collection device 103 or the microphone 204 mounted on each device.

  The voice analysis unit 302 analyzes voice information acquired by the voice acquisition unit 301 to acquire a voice command. The voice analysis unit 302 is realized by the control processing unit 212 of the control device 111. The voice analysis unit 302 analyzes the voice information acquired by the voice acquisition unit 301, and acquires a voice command indicating a control instruction to the device.

  The speaker identifying unit 303 identifies, based on the voice information acquired by the voice acquiring unit 301, the speaker information on the speaker who has uttered the acquired voice information. Here, the speaker specifying unit 303 specifies a rough speaker's age or gender. The speaker identification unit 303 is realized by, for example, the control processing unit 212 of the control device 111 and the database 213. The control processing unit 212 compares the voice data for each age or gender stored in the database 213 with the voice information acquired by the voice acquisition unit 301 to determine the age of the speaker who uttered the voice information or Estimate gender etc.

  That is, the speaker information includes information on the speaker's age. The database 213 stores a plurality of ages in association with voice models that can identify each age. In addition, the age may include not only one age such as 10 years old but also a predetermined age range such as 0-10 years old. The speaker identification unit 303 identifies a voice model that matches the voice information acquired by the voice acquisition unit 301, and identifies the age associated with the identified voice model.

  As the speaker information, not only rough information such as age or gender, but also the speaker may be specified in more detail. For example, the database 213 may store information on the detailed age, gender, preference, life pattern, etc. of each speaker in association with the voice information (voice model) of each speaker. In this case, the speaker specifying unit 303 can specify the speaker in more detail.

  The control target device specifying unit 304 specifies a control target device among the plurality of devices 101 based on the speaker information specified by the speaker specifying unit 303. The control target device identification unit 304 is realized by, for example, the control processing unit 212 of the control device 111 and the database 213. Here, in the present embodiment, the control target device specifying unit 304 specifies the control target device based on the speaker information, but the control target device specifying unit 304 is specified by the speaker specifying unit 303. The control target device may be specified based on only the speaker information, or the control target device may be specified based on information other than the speaker information specified by the speaker specifying unit 303. Details of the process of identifying the control target device based on information other than the speaker information will be described later.

  The device control unit 305 controls the control target device identified by the control target device identification unit 304 based on the voice command analyzed by the voice analysis unit 302. The device control unit 305 is realized by the control processing unit 202 of the device 101a.

(Flowchart of speech recognition system)
The flow of processing of the device control method in the present speech recognition system will be described below with reference to FIG.

  FIG. 7 is a flowchart showing the operation of the speech recognition system in the first embodiment of the present disclosure.

  First, in step S301, the speech acquisition unit 301 acquires speech information of a speaker.

  Next, in step S302, the voice analysis unit 302 analyzes the voice information acquired by the voice acquisition unit 301 to acquire a voice command.

  Next, in step S303, the speaker identifying unit 303 identifies, based on the voice information acquired by the voice acquiring unit 301, the speaker information on the speaker who has uttered the acquired voice information.

  Next, in step S304, the control target device identification unit 304 selects one of the plurality of devices 101 based on the voice command acquired by the voice analysis unit 302 and the speaker information identified by the speaker identification unit 303. Identify the control target device of Here, the control target device may be specified only from the specified speaker information.

  Next, in step S305, the device control unit 305 controls the control target device identified by the control target device identification unit 304.

(Operation example of speech recognition system)
An example of the flow of the operation of the speech recognition system according to the first embodiment will be described with reference to FIGS. 8 and 9. FIG. 8 is a sequence diagram for illustrating the operation of the first device of the speech recognition system in the first embodiment of the present disclosure. FIG. 9 is a sequence diagram for illustrating the operation of the second device of the speech recognition system in the first embodiment of the present disclosure. FIG. 10 is a diagram showing configurations of a control processing unit and a database in the first embodiment of the present disclosure.

  In the voice recognition system described below, as shown in FIG. 10, the control processing unit 212 of the control device 111 includes a voice analysis unit 302, a speaker identification unit 303, and a control target device identification unit 304. Further, as shown in FIG. 10, the database 213 of the control device 111 includes a voice command list 600 and a plurality of tables (table 700, table 801 and table 802).

  First, in step S501, the sound collection device 103 acquires audio information. Here, in the present speech recognition system, the timing of acquiring speech information is not limited. The sound collection device 103 may constantly acquire audio information, or may acquire audio information at a preset timing or interval. For example, the sound collection device 103 constantly acquires voice information in the installed space, and when detecting the frequency band of human voice in the acquired voice information, the acquired voice information is sent to the control device 111. You may output it.

  In addition, when the sound collection device 103 detects the frequency band of human voice in the acquired voice information, it detects a specific first word, and another second following the detected specific first word. May be acquired as voice information. For example, if the first word is "speech recognition", the speaker utters a second word following the first word, such as "speech recognition, temperature up". This makes it possible to identify whether the acquired voice information is a voice command. Note that the control device 111 may perform identification as to whether or not it is audio information, not the sound collection device 103.

  Here, the speech information acquired by the sound collection device 103 includes various noises in addition to the speech uttered by the speaker. Therefore, the sound collection device 103 may include a noise removal unit that removes noise included in the audio information. In addition, the control device 111 may include a noise removal unit, instead of the sound collection device 103 including a noise removal unit.

  The process of step S501 corresponds to the process of step S301 in FIG.

  Next, in step S502, the sound collection device 103 outputs the acquired voice information to the control device 111. Here, when outputting voice information, for example, in the configuration of the voice recognition system as shown in FIG. 1A, the sound collection device 103 directly connects the voice information by connecting the sound collection device 103 and the control device 111 by wire. You may output it. In addition, when the sound collection device 103 includes a communication unit for connecting to the network, the sound collection device 103 may transmit audio information to the control device 111 via the network. In addition, the sound collection device 103 may transmit voice information to the control device 111 via the home gateway 102. Further, the timing of outputting the audio information to the control device 111 is not particularly limited. For example, when the sound collection device 103 constantly acquires sound information, the sound collection device 103 records sound information for a certain period of time, and the sound collection device 103 records only when the frequency band of human voice is detected in the acquired sound information. The output voice information may be output to the control device 111.

  Next, in step S503, the voice analysis unit 302 of the control processing unit 212 of the control device 111 analyzes the obtained voice information to obtain a voice command. The speech analysis method may be a known method. As a method of speech analysis, for example, a method of analyzing acquired speech information, analyzing which speech command is close in the speech command list 600 registered in advance, and identifying the speech command can be considered. As the contents of the voice command list registered in advance, for example, an operation instruction to an apparatus such as "ON", "OFF", "GUE", and "PAG" can be considered. Here, as a result of the speech analysis unit 302 analyzing the speech information, if there is no word in the speech information that matches the registered speech command list 600, the process may not proceed to the next step S504.

  The process of step S503 corresponds to the process of step S302 in FIG.

  Next, in step S504, the speaker specifying unit 303 of the control processing unit 212 of the control device 111 specifies, from the acquired voice information, speaker information on the speaker who has uttered the voice information.

  Here, for example, the speaker specifying unit 303 of the control device 111 compares the acquired voice information with the voice information registered in the table 700 of the database 213 to estimate rough speaker information.

  FIG. 11 is a diagram showing an example of a table 700 showing the relationship between voice information recorded in the database 213 and the estimated age. As shown in FIG. 11, in the database 213, speech information (speech model) corresponding to the age range divided into five stages is registered. Here, the speech model refers to a discrimination model created using speech characteristics of an unspecified number of speakers corresponding to each age. The speech model contains features that indicate the speaker nature of speech of each age. For example, the speech model A is obtained by extracting features from an unspecified number of 0 to 10-year-old human speech, and learning and modeling the extracted features using some norm.

  The method of modeling the speech information is not particularly limited. Generally, it is known that the characteristics of the fundamental frequency or vocal tract of vocal cord vibration differ depending on gender or age, such as male, female, child and adult. Focusing on these differences, for example, a fundamental frequency extracted from speech information for each age or gender to be classified may be modeled, or feature quantities of vocal tract characteristics may be extracted, and the extracted feature fee may be machine learning or the like. It may be modeled by a statistical approach. Moreover, although the example which divided age into five steps and was database-ized was shown in this Embodiment, it does not restrict to this. Audio information corresponding to two age groups such as at least a child (for example, 0 to 10 years old) and other adults may be modeled and stored as a database. As in the table 700 of FIG. 11, voice information (voice model) and information related to age may be associated, or voice information (voice model) and gender or other information may be associated.

  The process of step 504 corresponds to the process of step S303 in FIG.

  Next, in step S505, the control target device identification unit 304 of the control processing unit 212 of the control device 111 identifies a control target device. Hereinafter, as an example of a method of specifying a control target device, a method of specifying a control target device from the specified speaker information (voice model) and the acquired voice command will be described. The method of specifying the control target device is not limited to the method described below. For example, a method of specifying the control target device only from the specified speaker information, or the specified speaker information and other information A method may be considered in combination to specify the control target device.

  As shown in FIG. 10, the control target device identification unit 304 may include a first control target device restriction unit 3041 and a second control target device restriction unit 3042.

  FIG. 12 is a diagram showing an example of a table 801 showing the relationship between the speech model stored in the database 213 and the control target device.

  The first control target device limiting unit 3041 limits control target devices from the speaker information (speech models A to D) specified by the speaker specifying unit 303 using the table 801 of the database 213 shown in FIG. . In FIG. 12, for example, the device A is a television, the device B is a gas stove, the device C is an air conditioner, and the device D is a toy for children. As shown in the table 801, when it is specified by the speaker specifying unit 303 that the age of the speaker is 0 to 10 years (voice model A), the control target devices are the device A (television) and the device D (children Limited to toys). When the speaker identification unit 303 identifies that the speaker age is 11 to 20 (voice model B), the control target devices are limited to the device A, the device C, and the device D. Similarly, when the speaker specifying unit 303 specifies that the age of the speaker is 21 to 40 (voice model C), the control target devices are limited to the device A, the device B, and the device C. The control target devices are limited as shown in FIG. 12 for other ages.

  Although the table 801 associates the voice model with the control target device, the present disclosure is not particularly limited thereto, and the age may be associated with the control target device.

  FIG. 13 is a diagram showing an example of a table 802 indicating the relationship between voice commands recorded in the database 213 and control target devices.

  The second control target device limiting unit 3042 limits the control target device from the voice command acquired by the voice analysis unit 302 using the table 802 of the database 213 shown in FIG. 13. For example, when the voice command is "on" or "off", the control target devices are limited to device A (TV), device B (gas stove), device C (air conditioner) and device D (child's toy) Ru. On the other hand, for example, when the voice command is "raising" or "teasing", the control target devices are limited to the device A (television), the device B (gas stove) and the device C (air conditioner).

  Here, for example, when the speaker identification unit 303 classifies the speech model A as the speech model A and the speech analysis unit 302 acquires a speech command “raising”, the control target device is the first control target device restriction unit 3041 and The device A (television) among the plurality of devices 101 limited by both the second control target device limiting unit 3042 is specified.

  As described above, the control target device specifying unit 304 associates the voice command with the candidate device by associating the table 801 (first table) storing the speaker information with the candidate device indicating the candidate of the control target device. Each of the tables 802 (second table) to be stored is referred to, and a candidate device that matches both the table 801 and the table 802 is specified as a control target device.

  In addition, when there is no candidate device that matches both of the table 801 (first table) and the table 802 (second table), the control target device identification unit 304 may notify that there is no control target device. .

  The process of step S505 corresponds to the process of step S304 in FIG.

  Next, in step S506, the communication unit 211 of the control device 111 transmits a voice command to the communication unit 201 of the control target device. As described above, when the device to be controlled is identified as the device A in step S505, the communication unit 211 transmits a voice command “raise” to the television as the device A as illustrated in FIG.

  On the other hand, when the control target device is specified as the device B in step S505, the communication unit 211 transmits a voice command to the communication unit 201 of the device B, as shown in FIG.

  Next, in step S507, the control processing unit 202 of the control target device controls the operation according to the received voice command.

  FIG. 14 is a diagram showing an example of a table 803 in which voice commands stored in the memory of the device and control contents are associated. Here, each device holds a table 803 as shown in FIG. For example, when a voice command “raise” is input to the device A, the control processing unit 202 controls the speaker 205 to raise the volume setting value by a fixed value.

  As described above, according to the voice recognition system of the first embodiment, it is possible to specify the control target device for the voice command of the speaker from the information such as the age of the speaker and the voice command. As a result, as in the conventional case, the input operation for specifying the control target device is not necessary, and the user's convenience can be further improved. In addition, it is possible to solve the problem of the conventional voice recognition system that anyone can operate all the devices by voice. For example, it is possible to prevent a child from operating equipment such as a gas stove by voice.

  In the first embodiment, an example in which the control processing unit 212 of the control device 111 includes the voice analysis unit 302, the speaker specifying unit 303, and the control target device specifying unit 304 has been described, but the present invention is not limited thereto. The control processing unit 202 in any one of the plurality of devices 101 may include any or all of the voice analysis unit 302, the speaker identification unit 303, and the control target device identification unit 304. Also, the server 121 owned by the service provider may include any or all of the voice analysis unit 302, the speaker identification unit 303, and the control target device identification unit 304. In that case, the data center operating company 110 accumulates the acquired information and transmits it to the service provider. Alternatively, the data center operating company 110 may not exist.

  The speaker information specified in step S303 may be the minimum rough information for determining whether to receive control of the device by voice, but the individual for each group is associated with the voice model in advance. If the table is registered, detailed speakers can be identified from the voice information. For example, the speaker can be specified in detail, such as whether the speaker is a father or a mother or a sister or a sister in a family. This makes it possible not to control the device even if speakers other than the previously registered speakers belonging to the group speak a voice command. Therefore, it is possible to prevent an outsider from controlling the devices in the group (such as a house).

  In addition, if a database for each group set in advance is accumulated, it is possible to count the frequency of use of the device for each speaker. Therefore, the control target device identification unit 304 may preferentially specify the device with high frequency of use of the speaker who utters the voice command as the control target device. That is, the database 213 may store a table that stores speaker information and the usage history of the device of the speaker corresponding to the speaker information and stores the table, and the control target device specifying unit 304 refers to the table. The control target device may be specified based on the usage history of the device of the speaker corresponding to the speaker information and the voice command. This makes it possible to prevent the control of the device not intended by the user. The details of the table associated with the individual for each group will be described in the second embodiment.

  Further, the control target device specifying unit 304 may specify the control target device using information on time when the device is used, together with the history of the use frequency. As a result, at the time when the speaker utters a voice command, it is possible to specify a device frequently used by the speaker as a control target device. Therefore, the accuracy of specifying the control target device can be further improved.

  Further, the control target device specifying unit 304 may specify the control target device by further utilizing the position information of the speaker and the device. That is, the control target device specifying unit 304 acquires speaker position information indicating the position of the speaker and device position information indicating the positions of the plurality of devices, and the voice command, the speaker information, and the speaker position information The control target device may be specified based on the device position information and the device position information.

  For example, after the control target device specifying unit 304 specifies a plurality of control target device candidates based on the voice command and the speaker information, the position of the speaker and the specified plurality of control target device candidates are identified. From the relationship with the position, the candidate for the control target device closest to the speaker position may be specified as the control target device. The position information may use, for example, a GPS (Global Positioning System) or the like. Further, based on the voice information acquired by the sound collection device 103, the positional relationship between the sound collection device 103 and each device and the positional relationship between the sound collection device 103 and the speaker may be measured. The plurality of devices 101 are provided with the microphone 204 and the speaker 205, and if the devices can exchange information, the positional relationship between the devices and the position of the speaker can be specified in more detail.

  Further, in step S305, when controlling the specified control target device, the control processing unit 202 may control the control target device according to the specified speaker. For example, when the speaker specified as voice model A (0 to 10 years old) in the table 700 controls the device A (TV) by a voice command, the display unit 206 of the device A The characters “let's leave 1 m or more” may be displayed, or the speaker 205 may output a voice “let's leave 1 m or more when watching TV”. In addition, when the speaker specified as the voice model E (60 to 80 years old) controls the device C by a voice command, the speaker 205 may output a more detailed voice guide regarding the operation method. That is, the control method of the device or the control result of the device in step S305 may be changed according to the information of the speaker.

  Further, in the tables shown in FIG. 12 and FIG. 13, the maker may set default in advance, or may receive editing by the user. In this case, it is preferable that a UI (user interface) for the user to edit each table is constructed.

  FIG. 15 is a view showing an example of a display screen for displaying a list of devices that can be operated by voice in a list, and FIG. 16 is a diagram for explaining an operation of setting a speaker operated by voice for each device. FIG. 17 is a diagram showing an example of a display screen displayed when a speaker operated by voice is set for each device.

  15 to 17 show UIs for the user to set each table. For example, as shown in FIG. 15, a list of devices that can be voice-operated is displayed on the screen of the smartphone. In the lower part of the screen, icons 501 and 502 for limiting the use of the device of a specific user are displayed. The icon 501 is an icon for restricting the operation by the 0 to 10-year-old speaker corresponding to the voice model A, and the icon 502 is for restricting the operation by the 61 to 80-year-old speaker corresponding to the voice model D It is an icon for The display unit for displaying the screen is configured of a touch panel display or the like. As shown in FIG. 16, the user drags and drops the icons 501 and 502 onto the target device. Thereby, the control processing unit 212 changes the control target device associated with the voice model corresponding to the icon in the table 801.

  In the example of FIG. 17, finally, the device B (gas stove) can not be operated by the voice command between the speaker of the voice model A (0 to 10 years old) and the speaker of the voice model D (61 to 80 years old) It is set to. Further, the device C (air conditioner) is set to a device in which the speaker of the voice model A (0 to 10 years old) can not operate by the voice command.

  As described above, the control processing unit 212 of the control device 111 controls a display device (for example, a smartphone) that displays information related to the plurality of devices 101 controlled based on the voice from the speaker. The control processing unit 212 controls the display device to display a list of first information related to a plurality of controllable audio devices. Further, the control processing unit 212 displays the second information indicating that the control of the device by voice is not received for the specific speaker so as to be displayed in the vicinity of the first information corresponding to the device not receiving the control. Control the device.

  Further, on the display screen of the display device, the first information and the second information arranged at a position different from the first information are displayed. In the control processing unit 212, the display area for displaying the second information on the display screen of the display device is touched by the user, the second information is dragged toward the first information, and the second information is the first information. By dropping on the information, the display device is controlled to display the second information in the vicinity of the first information. In addition, when the second information is displayed in the vicinity of the first information, the control processing unit 212 changes the setting for the specific speaker so as not to receive the control by the voice of the device corresponding to the first information. Do.

  The voice recognition system may also include an output unit that outputs the reason why the control target device identification unit 304 identifies the control target device. For example, in the case where the table is set as shown in FIGS. 12 and 13, when the speaker of the voice model E utters a voice command "Kyou", the control target device is automatically the device B (air conditioner (air conditioner) Specific to At this time, even if the communication unit 211 of the control device 111 transmits a control command to output a voice from the speaker 205 "The equipment that you can control by the voice command is the air conditioner, so control the air conditioner." Good. As a result, the speaker can automatically understand why the device B is controlled, thereby reducing discomfort to the speaker.

  In addition, the voice recognition system may include a control target device notification unit that visibly transmits the control target device to the speaker when the speaker is identified. For example, when a speaker present in a room in which the sound collection device 103 is installed is specified as the voice model A, candidates for the control target device are automatically specified as the device A and the device D. . At this time, the control device 111 may control the light emitting unit (not shown in FIG. 3) provided in the device A and the device D to emit light. As a result, the speaker can visually grasp the devices that can be controlled by the voice command.

  In addition, when the specified speaker tries to control a device that is not a control target device by a voice command, the control device 111 outputs a voice guide “the device is not a control target” from the speaker 104. You may control. For example, in the case where the tables of FIG. 12 and FIG. 13 are registered, even if the speaker specified as the speech model A utters “K”, the control target device does not exist. Therefore, when there is no control target device corresponding to the voice command, the communication unit 211 of the control device 111 transmits a control command indicating that the voice guide “the device is not controlled” is output from the speaker 205. It is also good.

Second Embodiment
(Findings that formed the basis of the second embodiment)
In the second embodiment, not only a plurality of devices are controlled by voice, but also a system that virtually interacts with each device or house by voice to purchase an item is assumed. For example, when the user tries to cook according to the recipe registered in the system, if the system side urges the purchase of the deficient food from the system side, and the user can instruct the purchase of the deficient food by voice. It is a convenient and valuable system for both users and service providers.

  However, at this time, for example, even a suspicious person who has invaded a house can purchase an article by voice input. Also, even young children can purchase expensive products. That is, anyone can purchase an item. Also, conventionally, when performing purchase settlement with a voice recognition system, a method for performing user authentication has not been established, so that the service provider is a user whose speaker who intends to purchase an article by voice is a user registered in advance. It can not be identified. Therefore, it is necessary to construct a system that can authenticate the user and purchase the item safely.

  The second embodiment provides a speech recognition system that solves the above-mentioned problems.

  In the second embodiment, a configuration different from that of the first embodiment will be mainly described. In the second embodiment, the configuration in the first embodiment can be used in combination. The overall configuration of the voice recognition system, the configuration of the device 101a, the configuration of the control device, and the like are the same as those in FIGS.

(Functional configuration of speech recognition system)
FIG. 18 is a diagram showing a functional configuration of the speech recognition system in the second embodiment of the present disclosure. The speech recognition system includes a speech acquisition unit 1001, a speech analysis unit 1002, a speaker identification unit 1003, and a purchase settlement unit 1004, as shown in FIG.

  The configurations of speech acquisition unit 1001, speech analysis unit 1002, and speaker identification unit 1003 are the same as the configurations of speech acquisition unit 301, speech analysis unit 302, and speaker identification unit 303 of the first embodiment, so the description will be omitted. . However, in the second embodiment, the speaker information specified by the speaker specifying unit 1003 is detailed speaker information. Detailed speaker information will be described later.

  The purchase settlement unit 1004 purchases and settles a product based on the voice command acquired by the voice analysis unit 1002 and the speaker information specified by the speaker specification unit 1003. The details of the purchase and settlement method will be described later. The purchase settlement unit 1004 is realized by the server 121 of the service provider 120. The detailed configuration of the server 121 will be described later.

  The purchase settlement unit 1004 may be realized by the control device 111 of the data center operating company 110.

(Flowchart of speech recognition system)
The flow of processing of the purchase / settlement method in the present speech recognition system will be described below using FIG.

  FIG. 19 is a flowchart showing an operation of the speech recognition system in the second embodiment of the present disclosure.

  First, in step S1101, the speech acquisition unit 1001 acquires speech information of the speaker.

  Next, in step S1102, the speech analysis unit 1002 analyzes the speech information acquired by the speech acquisition unit 1001 to acquire a speech command. Further, the voice analysis unit 1002 determines whether the voice command is a voice command related to purchase settlement.

  Next, in step S1103, the speaker specifying unit 1003 specifies the speaker information on the speaker who has uttered the obtained voice information based on the voice information obtained by the voice obtaining unit 1001. When the voice command is a voice command relating to purchase settlement, the speaker specifying unit 1003 specifies speaker information on a speaker who has uttered the obtained voice information based on the obtained voice information.

  Next, in step S1104, the purchase settlement unit 1004 performs purchase settlement of the product based on the voice command acquired by the speech analysis unit 1002 and the speaker information specified by the speaker specification unit 303. The purchase settlement unit 1004 refers to a table in which the purchase settlement possible speaker information is associated with the information necessary for the purchase settlement, and determines whether the identified speaker information is the purchase settlement possible speaker information. If it is determined that the identified speaker information is the speaker information that can be purchased and settled, the purchase settlement process is performed using the voice command and the information necessary for the purchase and settlement.

(Operation example of speech recognition system)
An example of the flow of the operation of the speech recognition system according to the second embodiment will be described with reference to FIG. FIG. 20 is a sequence diagram for illustrating purchase settlement processing of the speech recognition system according to the second embodiment of the present disclosure. FIG. 21 is a diagram showing configurations of a control device and a server in the second embodiment of the present disclosure.

  In the speech recognition system described below, as shown in FIG. 21, the control processing unit 212 of the control device 111 includes a speech analysis unit 1002 and a speaker identification unit 1003. Furthermore, the control processing unit 212 may include a control command selection unit 1006 and a person authentication unit 1008. Further, as shown in FIG. 21, the database 213 of the control device 111 includes a first voice command list 1301, a second voice command list 1302, a table associated with each group (house or company) (first table 1401 and a second table 1402).

  In addition, the server 121 of the service provider includes a communication unit 1210, a control processing unit 1220, and a purchase payment database 1230. Further, the control processing unit 1220 includes a purchase settlement unit 1004. Further, the control processing unit 1220 may include a registration presence / absence confirmation unit 1005. The purchase settlement database 1230 comprises a table 1403.

  First, in step S1201, the sound collection device 103 acquires audio information. Here, it is assumed that a voice command instructing purchase of a product is acquired. For example, voice information “purchase milk” is input to the sound collection device 103. The process of step S1201 corresponds to the process of step S1101 shown in FIG.

  Next, in step S1202, the sound collection device 103 transmits the acquired voice information to the control device 111.

  Next, in step S1203, the voice analysis unit 1002 of the control processing unit 212 of the control device 111 analyzes the obtained voice information to obtain a voice command. The voice analysis unit 1002 analyzes voice information as in the first embodiment. Here, the voice analysis unit 1002 checks whether the voice command includes a word intended for purchase settlement. Also, the voice analysis unit 1002 checks whether the voice command includes a word related to the product to be purchased.

  For example, when a voice command "purchase milk" is input, the voice analysis unit 1002 stores the word "milk" and the word "purchase" in advance, each of which stores the word meaning purchase. The voice command list 1301 is compared. In this case, since the word "purchase" exists in the first voice command list 1301, the voice analysis unit 1002 adds, as information, a voice command of purchase settlement. Further, the voice analysis unit 1002 compares the word “milk” with the second voice command list 1302 in which the names of articles that can be purchased are stored in advance. If the word "milk" is present in the second voice command list 1302, the voice analysis unit 1002 sets the product to be purchased as milk.

  The process of step S1203 corresponds to the process of step S1102 shown in FIG.

  Next, in step S1204, the speaker identification unit 1003 of the control processing unit 212 of the control device 111 identifies, from the acquired voice information, speaker information on the speaker who has uttered the voice information. The database 213 includes a table (first table 1401 and second table 1402) in which detailed information of members of each group is accumulated. The first table 1401 associates and stores voice information, age, and gender of members of the first group. The second table 1402 associates and stores voice information, age and gender of members of a second group different from the first group.

  FIG. 22 is a diagram showing an example of a first table 1401 showing the relationship between voice information recorded in the database 213, estimated age, and estimated sex. In the first table 1401, voice information, detailed age and gender are associated. Here, the audio information is as described in the first embodiment. The speaker identification unit 1003 identifies speaker information by comparing the voice information acquired in the process of step S1201 with the voice information registered in the first table 1401.

  The construction method of the first table 1401 is not particularly limited. For example, the control processing unit 212 may construct the first table 1401 by receiving in advance an initial setting using a UI (user interface) by the user. Further, the control processing unit 212 may construct the first table 1401 by causing the speaker to input information such as age and gender at any timing when the sound collection device 103 has acquired the voice information. In addition, when the speaker registers information in the purchase / settlement database 1230 described later, or after the speaker registers information in the purchase / settlement database 1230, the control processing unit 212 automatically performs purchase settlement on the first table 1401. Information of the database 1230 may be copied.

  The process of step S1204 corresponds to the process of step S1103 shown in FIG.

  Next, in step S 1205, when the communication unit 211 of the control apparatus 111 is determined as a voice command of purchase settlement as a result of analysis in step S 1203, the specified speaker information is used as the communication unit 1210 of the server 121. Send to The determination as to whether the voice command is a purchase payment voice command may not be made at this point. The communication unit 1210 of the server 121 receives the speaker information transmitted by the control device 111.

  Next, in step S1206, the registration presence / absence confirmation unit 1005 of the control processing unit 1220 of the server 121 confirms whether the speaker information identified in step S1204 is registered in the purchase settlement database 1230. The purchase payment database 1230 includes a table 1403 storing information on users who can purchase an item in the speech recognition system.

  FIG. 23 is a view showing an example of a table 1403 showing the relationship among user numbers, voice information, age and credit card information stored in the purchase payment database 1230.

  Here, when the voice information of the speaker identified in step S 1204 is present in the table 1403, the registration presence / absence confirmation unit 1005 stores the presence / absence of user registration as “presence”.

  For example, since the speaker specified as the voice information “001” in FIG. 22 does not exist in the table 1403 in step S 1204, the registration presence / absence confirmation unit 1005 stores the presence / absence of user registration as “absent”, and the step The process does not proceed to the process of S1207 and ends the process. The registration presence / absence confirmation unit 1005 may create a control command for outputting a voice guide such as “no registration information” or the like and transmit the control command to the speaker 104 without ending the process. On the other hand, for example, the speaker identified as voice information “003” in FIG. 22 in step S 1204 exists in the table 1403, so the registration presence / absence confirmation unit 1005 stores the presence / absence of user registration as “present”, and the step The process proceeds to the process of S1207.

  It should be noted that not only the match between the voice information in the table 1401 and the voice information in the table 1403 is determined when determining the presence or absence of user registration, but by using the match of other information (for example, age or gender). , The accuracy of authentication may be improved.

  Next, in step S1207, the communication unit 1210 of the server 121 transmits information indicating the presence or absence of user registration to the communication unit 211 of the control device 111. The communication unit 211 of the control device 111 receives the information indicating the presence or absence of the user registration transmitted by the server 121.

  Next, in step S1208, the control command selection unit 1006 of the control processing unit 212 of the control device 111 selects a control command for controlling the speaker 104 based on the voice command analyzed in step S1203. For example, if the word indicating the product matching the voice command list 1302 of the purchasable item in the voice command is an inexpensive product (here, milk), the control command selection unit 1006 performs “purchase settlement. Select (generate) a control command that outputs a voice guide saying "Are you sure you want to purchase?" In addition, if the word indicating the product that matches the voice command list 1302 is a relatively expensive product, the control command selection unit 1006 outputs a voice guide saying "Pay a purchase payment. Please speak a password". Select (generate) a control command.

  Here, although the control command selection unit 1006 changes the voice guide to be created depending on whether or not the product to be purchased is inexpensive, the present invention is not limited to this. The control command selection unit 1006 may always generate the same voice guide regardless of whether the product to be purchased is inexpensive. In addition, the determination as to whether or not the item to be purchased is inexpensive is determined by whether the item to be purchased is less than a predetermined price. The predetermined price may be set by the user.

  Next, in step S1209, the communication unit 211 of the control device 111 transmits the control command selected in step S1208 to the speaker 104. The speaker 104 receives the control command transmitted by the control device 111.

  Next, in step S1210, the speaker 104 outputs an audio guide based on the received control command. The voice guide is as described in step S1208. This can prompt the speaker to speak again.

  Next, in step S1211, the sound collection device 103 reacquires voice information of the speaker who is prompted to speak.

  Next, in step S1212, the sound collection device 103 transmits the acquired voice information to the control device 111.

  Next, in step S1213, the voice analysis unit 1002 of the control device 111 analyzes the voice again to obtain a voice command.

  Next, in step S1214, the speaker specifying unit 1003 of the control device 111 specifies, from the acquired voice information, speaker information on the speaker who has uttered the voice information.

  Next, in step S1215, the personal identification unit 1008 of the control processing unit 212 determines whether the speaker information identified in step S1204 matches the speaker information identified in step S1214, and the two stories If the user information matches, it is determined that the user authentication is successful.

  In step S1215, the personal authentication unit 1008 may determine whether the voice command analyzed in step S1213 is content that affirms purchase of a product. Here, if it is determined that the voice command is content that affirms the purchase of the product, the person authentication unit 1008 determines that the person authentication is successful. On the other hand, when it is determined that the voice command is content that denies the purchase of the product (for example, "No" or "Do not buy", etc.), the process may be ended. In this case, the control command selection unit 1006 may select (generate) a control command for outputting a voice guide of “end payment processing because there is no purchase intention”, and may transmit it to the speaker 104.

  Next, in step S1216, when it is determined in step S1215 that the authentication of the user is successful, the communication unit 211 of the control device 111 transmits, to the server 121, information indicating that the user authentication is successful. Here, if it is determined that the authentication of the user could not be performed, the control command selection unit 1006 selects (generates) a control command for outputting a voice guide saying "the authentication of the user can not be performed", and transmits it to the speaker 104. May be The communication unit 1210 of the server 121 receives the information indicating that the user authentication transmitted by the control device 111 is successful.

  Finally, in step S1217, the purchase settlement unit 1004 of the control processing unit 1220 of the server 121 performs purchase settlement processing for purchasing a product designated by voice. In the table 1403 of FIG. 23, the user number, the voice information, the age, and the credit card information are associated and managed. Therefore, in step S1217, the purchase settlement unit 1004 can perform purchase settlement processing by extracting the credit card information of the corresponding speaker and communicating with a card company or the like via the communication unit 1210.

  As described above, in the voice recognition system according to the second embodiment, whether the speaker is registered as a user by identifying the speaker from the acquired voice information and referring to the purchase / payment database stored in the server of the service provider It is determined whether or not there is a purchase settlement process. In this way, the user is not forced to perform troublesome authentication work. In addition, it is possible to prevent the danger of being able to purchase an item regardless of who speaks.

  In addition, identity verification is performed by confirming the purchase of the product again. This enables more secure purchase settlement processing.

  The control processing unit 212 may include a position detection unit that detects the current position of the speaker. After acquiring voice information in step S1201, the position detection unit detects the position of the speaker, and in step S1215, the position detection unit detects the position of the speaker again, and the personal identification unit 1008 detects 2 Authentication may be performed by confirming that the positions of the two speakers have not changed. As a result, even if a speaker different from the speaker from whom the voice information was acquired first is interrupted and uttered at the purchase payment decision timing (step S1211), the purchase payment can be canceled.

  The techniques described in the above aspects may be implemented, for example, in the following cloud service types. However, the type of cloud service in which the technology described in the above aspect is implemented is not limited to this.

(Service Type 1: In-house data center cloud service)
FIG. 24 is a diagram showing an overview of services provided by the speech recognition system in service type 1 (in-house data center type cloud service). In this type, the service provider 120 obtains information from the group 100 and provides a service to the user. In this type, the service provider 120 has the function of a data center operating company. That is, the service provider 120 has a cloud server (control device 111) that manages big data. Therefore, in this type, there is no data center operating company.

  In this type, the service provider 120 operates and manages the data center (cloud server) 2030. Also, the service provider 120 manages an operating system (OS) 2020 and an application 2010. The service provider 120 provides a service using the OS 2020 and the application 2010 managed by the service provider 120 (arrow 2040).

(Type 2 of service: Cloud service using IaaS)
FIG. 25 is a diagram showing an overview of services provided by the speech recognition system in service type 2 (IaaS-based cloud service). Here, IaaS is an abbreviation of infrastructure as a service, and is a cloud service providing model that provides the infrastructure itself for building and operating a computer system as a service via the Internet.

  In this type, the data center operating company 110 operates and manages the data center (cloud server) 2030. Also, the service provider 120 manages the OS 2020 and the application 2010. The service provider 120 provides a service using the OS 2020 and the application 2010 managed by the service provider 120 (arrow 2040).

(Type 3 of service: Cloud service using PaaS)
FIG. 26 is a diagram showing an overview of services provided by the speech recognition system in service type 3 (PaaS-based cloud service). Here, PaaS is an abbreviation of platform as a service, and is a cloud service providing model that provides a platform serving as a foundation for building and operating software as a service via the Internet.

  In this type, the data center operating company 110 manages the OS 2020, and operates and manages the data center (cloud server) 2030. Also, the service provider 120 manages the application 2010. The service provider 120 provides a service using the OS 2020 managed by the data center operating company 110 and the application 2010 managed by the service provider 120 (arrow 2040).

(Type 4 of service: Cloud service using SaaS)
FIG. 27 is a diagram showing an overview of a service provided by the speech recognition system in the service type 4 (SaaS cloud service). Here, SaaS is an abbreviation of software as a service. The SaaS-based cloud service is, for example, an application provided by a platform provider having a data center (cloud server), a user such as a company or an individual who does not have a data center (cloud server), the Internet, etc. Cloud service provision model having functions that can be used via the network of

  In this type, the data center operating company 110 manages the application 2010, manages the OS 2020, and operates and manages the data center (cloud server) 2030. The service provider 120 also provides a service using the OS 2020 and the application 2010 managed by the data center operating company 110 (arrow 2040).

  As described above, the service provider 120 provides a service in any type of cloud service. Also, for example, a service provider or a data center operating company may develop an OS, an application, a database of big data, etc. by oneself, or may outsource it to a third party.

  A device control method, a display control method, and a purchase settlement method according to the present disclosure can prevent a speaker who does not want to control a device to control the device by voice, and controls a plurality of devices by voice. The present invention is useful as a device control method, a display control method for controlling a display device that displays information on a plurality of devices controlled based on voice from a speaker, and a purchase settlement method for purchasing an article by voice.

10, 20 Users 100 Groups 101 Multiple Devices 101a Devices 102 Home Gateway 103 Sound Collection Device 104 Speaker 110 Data Center Operating Company 111 Control Device 120 Service Provider 121 Server 201 Communication Unit 202 Control Processing Unit 203 Memory 204 Microphone 205 Speaker 206 Display Unit 211 communication unit 212 control processing unit 213 database 301 speech acquisition unit 302 speech analysis unit 303 speaker identification unit 304 control target device identification unit 305 device control unit 1001 speech acquisition unit 1002 speech analysis unit 1003 speaker identification unit 1004 purchase settlement unit 1005 Registration existence check unit 1006 Control command selection unit 1008 Person authentication unit 1210 Communication unit 1220 Control processing unit 1230 Purchase settlement database 3041 First control target Instrument Limited 3042 second control target device limiting section

Claims (3)

It is a purchase settlement method for purchasing goods by voice,
Get voice information,
The acquired voice information is analyzed to obtain a voice command indicating a control instruction to the device,
When the voice command is a voice command related to purchase settlement, the speaker information on the speaker who utters the obtained voice information is specified based on the obtained voice information,
With reference to a table in which purchase settlement possible speaker information is associated with information necessary for purchase settlement, it is determined whether the identified speaker information is purchase settlement possible speaker information or not. When it is determined that the speaker information is purchasable-speaking speaker information, purchase settlement processing is performed using the voice command and the information necessary for the purchase settlement.
Purchase settlement method. If the price of the item to be purchased is equal to or less than a predetermined price, purchase approval processing is performed,
When the price of the article to be purchased is equal to or higher than a predetermined price, control is performed to output voice information prompting the user to utter a password for performing purchase decision processing.
The purchase decision method according to claim 1. Obtaining speaker position information on the position of the speaker;
Using the speaker position information, it is determined whether or not it is the utterance of the purchase-decided speaker.
The purchase decision method according to claim 1.