JP6543047B2 - Information processing apparatus, control program, recording medium - Google Patents

Description

  The present invention relates to an information processing apparatus that presents an instruction to a subject so that the subject performs a specific operation, a control method of the information processing apparatus, a program, and a recording medium.

  A robot apparatus and the like capable of registering a person's name and a face with the other party of the dialogue as a recognition target using a dialogue function with a human is known as a prior art.

  For example, Patent Document 1 searches data by face photograph, person name and voice feature, and if the voice of the speaker is new, this speaker is a recognition target, and that person name and face are registered. A robot apparatus is described that instructs a user to perform a specific action such as speaking a favorite food after emitting a voice to a recognition object and giving a name.

Unexamined-Japanese-Patent No. 2003-255989 (Sep. 10, 2003 publication)

  However, since the robot apparatus described in Patent Document 1 determines whether the speaker is a new recognition target based on whether the speaker's voice is new or not, the case where the speaker's voice is not new There is a problem that even if the voice contains the name of another person, that person is not to be recognized.

  For example, when a non-new (that is, a recognized) speaker introduces a new recognition target or speaks to the recognition target, the technology described in Patent Document 1 allows the speaker's voice to be new. Because there is no such person, it may not be possible to correctly determine whether the person introduced or spoken by the speaker is a new recognition target or a recognized person. And if this judgment is wrong, it is not possible to recognize a new recognition target person's name.

  The present invention has been made in view of the above points, and an object thereof is to instruct a subject to perform a predetermined action when a voice including the person's name of the subject is detected. It is to realize an information processing apparatus capable of processing information and a control method of the information processing apparatus.

  In order to solve the above-mentioned subject, an information processor concerning one mode of the present invention is an information processor provided with a voice input part which receives an input of a voice, and it is an object person's person who was inputted from the above-mentioned voice input part. The identification information extraction unit extracts an identification information extraction unit that extracts identification information of the target person from voice data of voice including identification information, and an operation instruction that instructs the target person to perform a predetermined operation. And an instruction presenting unit that presents the operation instruction to the subject.

  Further, in order to solve the above problems, a control method of an information processing apparatus according to an aspect of the present invention is a control method of an information processing apparatus including a voice input unit that receives an input of voice, the voice input unit A voice input step for receiving an input of voice including identification information of a target person, an identification information extraction step for extracting identification information of the target person from voice data of the voice received in the voice input step, and An instruction generation step of generating an operation instruction instructing the person to perform a predetermined operation including the identification information extracted in the identification information extraction step; and the operation instruction generated in the instruction generation step And D. an instruction presenting step presented to the subject.

  According to one aspect of the present invention, when a voice including a person's name of a subject is detected, there is an effect that the subject can be instructed to perform a specific action.

It is a block diagram which shows an example of schematic structure of the information processing apparatus which concerns on this invention. It is a block diagram showing an example of the hardware constitutions of an information processor. FIG. 7 is a diagram for explaining a process in which the information processing apparatus registers a person name of a registration target person in association with feature data; FIG. 6 is a diagram illustrating an example of processing of extracting text of a person's name from text data by the information processing apparatus. It is a figure which shows an example of the template of the audio | voice which an information processing apparatus outputs. 5 is a flowchart illustrating an example of the flow of processing in which an information processing apparatus registers a person name of a registration target person and feature data in association with each other;

Embodiment 1
Hereinafter, embodiments of the present invention will be described in detail.

(Process in which the information processing apparatus 1 registers a registration target person)
First, in one embodiment of the present invention, a process of associating and registering the person name (identification information) of a registration target person (target person) to be registered by the information processing apparatus 1 and the feature data (specific information) This will be described with reference to FIG. FIG. 3 is a diagram for explaining processing in which the information processing apparatus 1 registers the person name of the registration target person and the feature data in association with each other.

  The information processing apparatus 1 receives an input of a voice including a person's name (first name, last name, etc.) of a person to be registered, extracts the person's name of the person to be registered included in the voice, and selects the person's name of the person to be registered. A voice instruction is issued to instruct the person to be registered to perform a predetermined operation.

  Extraction of the person's name of the registration subject may be when the input of the voice emitted by the registration subject himself / herself may be received, or when the input of the voice emitted by a person (such as the user U) different from the registration subject is accepted. It may be Here, as an example of the voice emitted by the person to be registered, there is voice when the person to be registered introduces himself or gives a greeting by giving his / her name to another person. On the other hand, as an example of a voice emitted by a person different from the registration subject, a voice when the user U introduces the registration subject to another person, when the user U introduces the registration subject to the information processing apparatus 1 And the voice when the user U asks or speaks with the person to be registered. If the voice of the person to be registered or the voice of a person different from the person to be registered includes the person's name of the person to be registered, such voice has a sentence pattern of a predetermined format. Further, the person name of the registration target person extracted by the information processing apparatus 1 is not limited to a person name such as a surname or a first name, as long as the registration target person is a name recognized as a name pointing to himself. (Common name), another name, stage name, etc. may be used.

  In the following, the face of the registration subject who has performed a predetermined action according to the instruction is imaged, feature data is extracted from the face image F obtained by imaging the face of the registration subject, and the feature data and the person name of the registration subject The information processing apparatus 1 that generates and manages registration data in which the above are associated will be described as an example. Here, the “feature data” of the registration subject is not limited to the feature data extracted from the face image F of the registration subject, but is the feature data extracted from fingerprint information, retina information, voice print information, etc. It is also good.

  In the following, the case where the “predetermined action” is an instruction to bring the face of the registration target person closer to the face of the registration target person to obtain the face image F obtained by imaging the face of the registration target person is described as an example. Although it is not limited to this. That is, although the information processing apparatus 1 instructs the registration target person to perform a predetermined operation, the “predetermined operation” is because the information processing apparatus 1 acquires unique information unique to the registration target person. The operation is not limited to the above and may be any operation. For example, in the case of an instruction for acquiring the fingerprint information of the registration target person, the information processing apparatus 1 may instruct that “adhere the finger closely to the scanner surface” as “predetermined operation”. Alternatively, it may be instructed that “the registration target person causes the information processing apparatus 1 to perform predetermined operation input” or “moves to the predetermined position” as “predetermined operation”. In such a case, the “person to be registered” is not limited to the person to be registered in the information processing apparatus 1, and the person to be instructed to perform a predetermined operation according to the instruction issued by the information processing apparatus 1 (object person ) May be.

  For example, as shown in (a) of FIG. 3, the information processing apparatus 1 includes a voice input unit 31 and a voice output unit 32 (instruction presentation unit) for receiving voice input, and together with Mr. Suzuki and Mr. Sato A voice of “This is Mr. Suzuki” issued by the user U who is present in is acquired. Then, the information processing apparatus 1 uses the “Suzuki”, which is the personal name of the person to be registered included in the voice uttered by the user U, and instructs the Mr. Suzuki to bring his face closer “Mr. Suzuki, "Please bring the face close" from the voice output unit 32.

  The information processing device 1 further includes an imaging device 33, and captures an image of a face brought close after outputting a voice instruction. For example, as shown in (b) of FIG. 3, when Mr. Suzuki who has heard the voice instruction approaches the information processing apparatus 1 and brings the face closer, the information processing apparatus 1 captures an image of the face and The face image F as shown in) is acquired.

  Thus, the information processing apparatus 1 sends a voice instruction including the name of the person to be registered (for example, “Suzuki”) included in the voice uttered by the user U to the person to be registered with the person name. Make a predetermined action, such as bringing the face closer. In the example shown in (a) of FIG. 3, although a plurality of persons (that is, the user U, Mr. Suzuki and Mr. Sato) listen to the voice instruction from the information processing apparatus 1, the person to be registered in the voice instruction By calling and including the person's name, it is possible to make a specific person to be registered perform a predetermined operation. By bringing the face closer to the person of the person name (for example, Mr. Suzuki) included in the voice instruction, the face of the person to be registered can be correctly imaged without being confused with another person.

  The information processing apparatus 1 extracts one or a plurality of feature data from a face image F obtained by capturing the face of Mr. Suzuki, who brings the face close according to a voice instruction, and extracts the extracted feature data and the person name of the registration target person "Mr. Suzuki" is associated with each other to generate registration data, and this registration data is stored in the registration database 24 (see FIG. 1). Extraction of feature data from the face image F is performed, for example, by the positions of the points P1 and P2 corresponding to the eyes of the face image F, the position of the point P3 corresponding to the nose, and the point P4 corresponding to the mouth (center of the mouth) , Position coordinates of each position of point P5 (right end of the mouth) and point P6 (left end of the mouth), and calculating the distance between each point. Here, the position coordinate is a coordinate corresponding to the position of each point in a coordinate system whose origin is a predetermined position (for example, the position of the lower left end) of the face image F. The feature data to be extracted is not limited to these, and arbitrary feature data such as image data of the face image F itself, face color, shape of ear, etc. may be used as registration data. Further, the face image F is not limited to a planar image, and may be a stereoscopic image.

  Although the information processing apparatus 1 includes the imaging device 33 and issues an instruction voice (operation instruction) to bring the face closer in order to capture an image of the face of the registration target person, the instruction voice has been described. Is not limited to this. For example, after imaging the face of Mr. Suzuki, further issue a voice instruction “Please show Mr. Suzuki, please show your ears.” And also imaging Mr. Suzuki's ears, adding the shape of the ears to the feature data. It is also good. In addition, when the information processing apparatus 1 includes a fingerprint sensor, a voice instruction of "Ms. Suzuki, put a finger in contact with the scanner surface" may be issued to acquire Mr. Suzuki's fingerprint information as feature data.

  In addition, although an example in which an operation instruction for capturing a face of a registration target person is issued as a voice has been described, the output of the operation instruction is not limited to the voice output, and a predetermined operation is performed for the registration target person As long as it is an output method that can be presented as such, it may be any method. For example, the operation instruction may be displayed as a character string on a display device (not shown).

(Hardware configuration of the information processing apparatus 1)
First, the hardware configuration of the information processing device 1 will be described with reference to FIG. FIG. 2 is a block diagram showing an example of the hardware configuration of the information processing apparatus 1. Note that for convenience of description, the description of other functions that can be included in the information processing apparatus 1, such as an operation control unit that controls attitude and the like and a communication unit that performs data communication, will be omitted.

  The information processing apparatus 1 includes an audio input unit 31, an audio output unit 32, an imaging device 33, a control unit 10, and a storage unit 20. The control unit 10 performs control to execute the processing of each function of the information processing apparatus 1, and acquires the voice input from the voice input unit 31 and the captured image R captured by the imaging device 33. Further, the control unit 10, the recognition dictionary 21 stored in the storage unit 20, the morphological analysis dictionary 22, the instruction template database 23 (see FIG. 1), and the like are appropriately referenced to execute predetermined processing. Further, the control unit 10 stores the generated registration data in the registration database 24.

(Configuration of information processing apparatus 1)
Next, the configuration of the information processing apparatus 1 will be described with reference to FIG. FIG. 1 is a block diagram showing an example of a schematic configuration of the information processing apparatus 1. The information processing apparatus 1 includes the voice input unit 31, the voice output unit 32, the imaging device 33, the control unit 10, and the storage unit 20, as already described in FIG. The control unit 10 that controls each function included in the information processing apparatus 1 further includes a textification unit 11, a morphological analysis unit 12, an introduction sentence extraction unit 13, a person name extraction unit 14 (identification information extraction unit), and a voice instruction generation unit The image processing apparatus includes a command generation unit 15, an imaging device control unit 16, a feature data extraction unit 17 (specific information extraction unit), and a registration data management unit 18. The storage unit 20 includes a recognition dictionary 21, a morphological analysis dictionary 22, an instruction template database 23, and a registration database 24 (storage unit).

  The voice input unit 31 is a microphone for receiving voice input, and voice data of the voice input from the voice input unit 31 is sent to the text conversion unit 11.

  The audio output unit 32 is a speaker that outputs audio data as audio. The voice output unit 32 outputs the voice data of the voice instruction acquired from the voice instruction generation unit 15.

  The imaging device 33 is a digital camera that acquires image data, and is, for example, a camera module that captures an image of an imaging target using a CCD sensor, a CMOS sensor, or the like. The image captured by the imaging device 33 is not limited to a two-dimensional image, and may be a three-dimensional image.

  The text conversion unit 11 calculates an acoustic feature of voice data of voice input from the voice input unit 31, and converts the voice data into text data while referring to the recognition dictionary 21. An acoustic model and a language model are defined in the recognition dictionary 21, and the textification unit 11 assigns text corresponding to each acoustic feature amount, and converts voice data into text data. Specifically, the recognition dictionary 21 represents what kind of frequency characteristic the speech included in the speech data has, and corresponds to a large number of speech waveform samples and corresponding text data. Relationships are defined. The text conversion unit 11 searches the recognition dictionary 21 and specifies the text corresponding to the voice that matches the calculated acoustic feature amount. The text conversion unit 11 sends the text data to the morphological analysis unit 12.

  The morphological analysis unit 12 decomposes the text data acquired from the text conversion unit 11 into morphemes. A morpheme is the smallest linguistic unit having meaning, and is a component showing a grammatical relation. The morphological analysis unit 12 decomposes the text data into constituent elements on the grammar with reference to the morphological analysis dictionary 22, and specifies the grammatical relationship of the constituent elements. The morphological analysis unit 12 outputs the analysis result to the introductory sentence extraction unit 13 together with the text data to be analyzed. A specific example of morpheme analysis in text data in which the morpheme analysis unit 12 converts the voice of the user U introducing the registration target into text will be described later.

  The introductory sentence extraction unit 13 extracts a sentence including the person's name of the registration target person from the text data (text) generated by speech recognition based on the result of the morphological analysis. The introduction sentence extraction unit 13 introduces a self-introduction to a person different from the person to be registered or a person to be registered introduces the person to be registered to a third party or the information processing apparatus 1. The introduction sentence etc. may be extracted from the voice of. For example, an introduction sentence such as “I am Suzuki.” Who is the subject of the self-introduction, or a person (eg, user U) who is different from the subject of the registration such as “This is Mr. Suzuki” (the user U) The introductory sentence to introduce has a prescribed form that is characteristic of the introductory sentence, such as the person name “Suzuki” is placed after the first person or third person pronoun (I, I, I, here, etc.) There is. The form of the introduction including the person's name of the person to be registered is not limited to this. For example, Mr. Suzuki, who is the subject of registration, emits the first-person pronoun when he / she introduces himself to user U and Mr. Sato, and emits "I am Suzuki. Thank you all," and "Mr. Sato, nice to meet you, Suzuki. The text data generated from the voice such as “.” Also has a sentence pattern in a predetermined format including the person's name of the registration target, and may be extracted as an introduction sentence. The introduction sentence extracting unit 13 extracts text data having a sentence pattern of a predetermined format including the person's name of the registration target.

  The person name extraction unit 14 determines the noun specified by the morphological analysis unit 12 as “noun, proper noun, person's name, surname” in the text data confirmed to be the introduction sentence by the introduction sentence extraction unit 13, Extract as text data of person's name. The text data of the extracted personal name is output to the voice instruction generation unit 15 and the registration data management unit 18.

  The voice instruction generation unit 15 acquires the text of the person name (for example, “Suzuki”) extracted by the person name extraction unit 14, and the person name is acquired at a predetermined position of the instruction template stored in the instruction template database 23. Is inserted to generate a voice instruction including the person's name. Note that a specific example of processing in which the voice instruction generation unit 15 generates a voice instruction including the personal name of the registration target person introduced by the user U will be described later.

  The imaging device control unit 16 controls the imaging device 33 to capture an image when the registration target person who has heard the voice instruction output from the voice output unit 32 performs a predetermined operation instructed by the voice instruction. In addition, when text data of a person's name extracted by the person's name extraction unit 14 is output from the person's name extraction unit 14 to the voice instruction generation unit 15, the imaging device control unit 16 starts imaging by the imaging device 33. You may instruct to Alternatively, when the voice instruction generation unit 15 outputs the generated voice instruction to the voice output unit 32, the imaging device control unit 16 may be instructed to start imaging by the imaging device 33.

  In addition, the imaging device control unit 16 determines whether the captured image R captured by the imaging device 33 satisfies the condition to be satisfied for extracting the feature data, and in the case where the condition is not satisfied. The voice instruction generation unit 15 may be instructed to output the same voice instruction or a voice instruction (related voice instruction) instructing a related operation. The conditions to be satisfied in order to extract the feature data include, for example, that the area of the face image F in the captured image R is a certain size or more, each position of the eyes, nose, and mouth in the captured face image F The position coordinates of points P1 to P6 in (c) of FIG. 3 can be determined.

  The feature data extraction unit 17 detects the positions of the points P1 and P2 corresponding to the eyes of the face image F, the position of the point P3 corresponding to the nose, and the points P4 (center of the mouth) and P5 (for the mouth). The position coordinates of each position of the right end) and the point P6 (left end of the mouth), the distance between each point, and the like are extracted as feature data.

  The registration data management unit 18 associates the feature data extracted by the feature data extraction unit 17 with the person name to generate registration data, and stores the registration data in the registration database 24.

  As described above, when the information processing apparatus 1 receives an input of the voice of the user U who introduces the registration target person to be registered, the information processing device 1 includes the personal name of the registration target person so that the registration target person performs a predetermined operation. Give voice instructions. As a result, the face of the person to be registered can be imaged without being confused with another person, so that registration data regarding the person to be registered can be accurately managed.

  The voice targeted for extracting the person name of the registration target person is a voice where the user U introduces the registration target person to a third party or the information processing apparatus 1, or the registration target person introduces himself to the other person The voice is not limited to the voice, but may be voice when the user U speaks or asks the person to be registered. For example, in the case shown in FIG. 3, the user U asks Mr. Suzuki or Mr. Sato a voice asking "Suzule, how are you?" Or a voice talking "Sato, please go here". It may be. In such a case, the introductory sentence extraction unit 13 (target sentence extraction unit) generates a predetermined format to be emitted to a person other than the person from text data generated by speech recognition based on the result of morphological analysis. Extract text data (target sentence) of speech having sentence pattern. Next, in the text data in which it is confirmed by the introduction sentence extraction unit 13 that the person name extraction unit 14 has a sentence pattern in a predetermined format, the morphological analysis unit 12 is “noun, proper noun, person's name, surname” The noun identified as can be extracted as text data of a person's name.

(Extraction of person's name from text data)
Next, a process of the person name extraction unit 14 extracting a person name included in text data based on morpheme analysis of text data by the morpheme analysis unit 12 will be described using FIG. 4. FIG. 4 is a diagram showing an example of processing for extracting a text of a person's name from text data. Figure 4 (a) explains the morphological analysis of the text data "This is Mr. Suzuki." And extraction of the person's name, and (b) the text "This is Mr. Suzuki and Mr. Sato." It explains morpheme analysis of data and extraction of a person's name.

  As shown in (a) of FIG. 4, the text data “This is Mr. Suzuki.” Is decomposed into morphemes “here”, “ha”, “suzuki”, “san”, and “is”. . "Here" is attributed to "noun, pronoun, general (noun)", "ha" is attributed to "article, associated particle", and "Suzuki" is associated with "noun, proper noun, person" The family name is surnamed, the "san" is attributed to the "noun, suffix, person's name", and the "is" is attributed to the "auxiliary verb". The person's name extraction unit 14 extracts text data "Suzuki" to which the morphological analysis unit 12 belongs as "noun, proper noun, person's name, surname" as a person's name.

  As shown in (b) of FIG. 4, the text data “This is Mr. Suzuki and Mr. Sato” is “here”, “Ha”, “Suzuki”, “Mr”, “To”, “Sato” It is decomposed into morphemes, "san", and "is". Then, "Suzuki" and "Sato" are attributed to "Noun, proper noun, person's name, surname". The person's name extraction unit 14 extracts text data "Suzuki" to which the morphological analysis unit 12 belongs as "noun, proper noun, person's name, surname" as a person's name.

  As described above, the personal name extraction unit 14 may extract text data of one or more personal names from each text data. Thus, even when a voice introducing a plurality of persons is input from the user U, the personal name extraction unit 14 can extract all the personal names of the registration target persons introduced by the voice.

(Generation of sound to be output)
Subsequently, an example of a voice instruction to instruct a registration subject so that the registration subject performs a predetermined operation will be described with reference to FIG. FIG. 5 is a diagram showing an example of a sound template output by the information processing apparatus 1. The voice instruction generation unit 15 reads out the instruction template from the instruction template database 23, inserts the person name acquired from the person name extraction unit 14 into each instruction template, and generates a voice instruction.

  The voice instruction "Mr. Suzuki, please bring your face close" outputted from the information processing apparatus 1 shown in (a) of FIG. 3 is the instruction template "(person's name) Mr. face of the voice instruction A1. It can be generated by inserting “Suzuki”, which is a person's name acquired from the person's name extraction unit 14, in the “(person's name)” portion of “Please move closer”.

  Note that the imaging device control unit 16 determines whether the captured image R captured by the imaging device 33 satisfies a condition to be satisfied for extracting feature data, and if this condition is not satisfied. The voice instruction generation unit 15 may be instructed to output the same voice instruction or a voice instruction (related voice instruction) instructing a related operation, and imaging of the face image may be redone.

  For example, when it is determined that the condition for extracting feature data is not satisfied because the area of the face image F in the captured image R is not equal to or larger than a certain size, the imaging device control unit 16 By inserting the person name of the registration target person acquired from the person name extraction unit 14 in the part of “(person name)” of the instruction template “(person name), please bring your face a little closer”. A voice instruction may be generated to make the person to be registered more closely approach the face. Similarly, the person of the registration target person acquired from the person name extraction unit 14 in the part of “(person name)” of the instruction template “(person name) Mr. face please face here” of the voice instruction A3. By inserting the name, a voice instruction may be generated to make the person to be registered more closely approach the face.

  Before generating the voice instruction, an auxiliary phrase sound as shown in (b) of FIG. 5 may be generated and output. In the auxiliary phrase voice, for example, the person name of the registration target person acquired from the person name extraction unit 14 is inserted in the part of “(person name)” of the assistance phrase X1 template “(person name) Mr. nice to meet you.” Generated by By outputting such a voice, it is possible to urge a person to be registered to prepare to listen to the voice instruction, and to increase interest in the voice instruction to be output thereafter.

  Furthermore, when the imaging of the face is successful, the auxiliary phrase sound as shown in (b) of FIG. 5 may be generated and output. The auxiliary phrase sound is, for example, the auxiliary phrase X2 template "(person name) Mr. Thank you very much.", The auxiliary phrase X3 template "(person name) Mr. thank you." It is generated by inserting the person name of the registration target person acquired from the name extraction unit 14. By outputting such a voice, the user U who introduced the registration target person that imaging of the face of the registration target person is completed, that is, proceeding to processing for extracting feature data and generating registration data And the person to be registered can be notified.

(Flow of processing in which the information processing apparatus 1 registers the person name of the registration target person and the feature data in association with each other)
Here, the flow of processing in which the information processing apparatus 1 registers the person name of the registration target person in association with the feature data will be described with reference to FIG. FIG. 6 is a flow chart for explaining an example of the flow of processing in which the information processing apparatus 1 registers the person name of the registration target person in association with the feature data.

  First, the information processing apparatus 1 starts accepting voice input from the voice input unit 31, and receives voice input from the user U for introducing a person to be registered (voice input step). The text conversion unit 11 calculates the acoustic feature amount of the voice data of the input voice (S1). Next, the text conversion unit 11 converts the voice data of the input voice into text data based on the correspondence between the acoustic feature amount and the text defined in the recognition dictionary 21 (S2). Subsequently, the morphological analysis unit 12 decomposes the text data into each component in the grammar, and refers to the morphological analysis dictionary 22 to specify the grammatical relationship of each component (S3). When the introduction sentence extraction unit 13 confirms that the introduction sentence introduces the person to be registered from the result of morphological analysis of the text data (S4), the person name extraction unit 14 confirms that the introduction sentence is In the text data, the morphological analysis unit 12 extracts words (nouns) belonging as, for example, "noun, proper noun, person's name, surname" from text data as person's name (S5: identification information extraction step). The person name extraction unit 14 also sends the extracted person name to the voice instruction generation unit 15 and the registration data management unit 18.

  The voice instruction generation unit 15 having acquired the text of the person's name reads out the instruction template from the instruction template database 23, inserts the person's name acquired from the person's name extraction unit 14 into each instruction template, and the registration target person is predetermined. A voice instruction instructing operation is generated (S6: instruction generation step). Here, the voice instruction generation unit 15 sends an instruction to the imaging device control unit 16 to start imaging with the imaging device 33 before generating the voice instruction or immediately after generating the voice instruction. The imaging by the device 33 starts (S7). Thereafter, an audio instruction is output from the audio output unit 32 (S8: instruction presentation step). In addition, the order of the process of S6-S8 is an example, and is not limited to this. For example, imaging may be started by the imaging device 33 immediately after the audio instruction is output from the audio output unit 32.

  Since the voice instruction output by the information processing apparatus 1 includes the name of the person to be registered, the person to be registered calls and instructs only the person to be registered so as to perform a predetermined operation. For example, as shown in (a) of FIG. 3, Mr. Suzuki who is a registration target who has heard the voice instruction "Ms. Suzuki, please bring your face close" brings the face close to the information processing apparatus 1, but the user Since U and Mr. Sato have not received an instruction, the face can not be brought close to the information processing apparatus 1. The information processing apparatus 1 acquires a captured image R obtained by capturing an approached face as a face image F of Mr. Suzuki, who is a registration target.

  The imaging device control unit 16 detects a face image F from the captured image R (S9), and determines whether a condition to be satisfied for extracting feature data is satisfied (S10). FIG. 6 exemplifies a case where the condition to be satisfied for extracting the feature data is that the area of the face image F in the captured image R is a size equal to or larger than the threshold. If this condition is not satisfied (NO in S10), the process returns to S8, and the voice instruction generation unit 15 is instructed to output the same voice instruction or a voice instruction (related voice instruction) instructing the related operation. Do. On the other hand, if the feature data can be extracted (YES in S10), the captured image R is sent from the imaging device control unit 16 to the feature data extraction unit 17, and the feature data extraction unit 17 extracts feature data of the face image F (S11).

  Finally, the registration data management unit 18 stores in the registration database 24 the registration data of the person to be registered, in which the feature data acquired from the feature data extraction unit 17 is associated with the person name acquired from the person name extraction unit 14 (S12: registration data management step).

  As described above, when the information processing apparatus 1 receives an input of a voice introducing a registration target person to be registered, a voice instruction including the person name of the registration target person is performed so that the registration target person performs a predetermined operation. It emits. As a result, the face of the person to be registered can be imaged without being confused with another person, so that registration data regarding the person to be registered can be accurately managed.

Second Embodiment
In the above-described example, the information processing apparatus 1 includes the voice input unit 31, the voice output unit 32, and the imaging device 33, and generates and manages registration data regarding a registration target person introduced by the user U. Not limited to this. For example, if it is possible to transmit and receive data between the information processing device 1, the voice input unit 31, the voice output unit 32, and the imaging device 33, the information processing device 1, the voice input unit 31, the voice output unit 32, And the imaging device 33 may be configured separately.

  In this case, the voice data of the voice input to the voice input unit 31 is transmitted to the information processing device 1. The information processing apparatus 1 converts the received voice data into text data, performs morphological analysis of the text data, and extracts the person's name of the registration target person included in the text data. The information processing apparatus 1 transmits voice data of a voice instruction generated using the person's name to the voice output unit 32, and transmits an instruction to start imaging to the imaging device 33.

  The captured image R captured by the imaging device 33 is transmitted to the information processing device 1, and the information processing device 1 extracts feature data of the registration target person from the face image F of the captured image R, and the person of the registration target person Registration data is generated in association with a name and stored in the storage unit 20.

  As described above, the voice input unit 31, the voice output unit 32, and the imaging device 33 can also be installed at a position apart from the position at which the control unit 10 of the information processing device 1 is installed.

Third Embodiment
Control blocks of the information processing apparatus 1 (in particular, the text conversion unit 11, the morphological analysis unit 12, the introduction sentence extraction unit 13, the person name extraction unit 14, the voice instruction generation unit 15, the imaging device control unit 16, the feature data extraction unit 17, The registration data management unit 18) may be realized by a logic circuit (hardware) formed in an integrated circuit (IC chip) or the like, or may be realized by software using a CPU (central processing unit) .

  In the latter case, the information processing apparatus 1 is a CPU that executes instructions of a program that is software that realizes each function, and a ROM (Read Only Memory) in which the program and various data are readably recorded by a computer (or CPU). Alternatively, a storage device (these are referred to as “recording media”), a RAM (Random Access Memory) for developing the program, and the like are provided. The object of the present invention is achieved by the computer (or CPU) reading the program from the recording medium and executing the program. As the recording medium, a “non-transitory tangible medium”, for example, a tape, a disk, a card, a semiconductor memory, a programmable logic circuit or the like can be used. The program may be supplied to the computer via any transmission medium (communication network, broadcast wave, etc.) capable of transmitting the program. The present invention can also be realized in the form of a data signal embedded in a carrier wave, in which the program is embodied by electronic transmission.

[Summary]
The information processing apparatus 1 according to aspect 1 of the present invention is an information processing apparatus including an audio input unit 31 that receives an input of audio, and is an audio of audio including identification information of a target person input from the audio input unit. The identification information extraction unit extracts an identification information extraction unit (person name extraction unit 14) that extracts identification information of the target person from data, and an operation instruction that instructs the target person to perform a predetermined operation. It includes an instruction generation unit (voice instruction generation unit 15) that generates the identification information and the above, and an instruction presentation unit (voice output unit 32) that presents the operation instruction to the target person.

  According to the above configuration, the identification information of the target person is extracted from the voice including the identification information of the target person, and the operation instruction including the identification information of the target person is performed so that the target person performs a predetermined operation. Present to the above subjects. In this way, it is possible to cause the correct subject to perform a predetermined action without being confused with another person.

  In the information processing device according to aspect 2 of the present invention, in the aspect 1, the sound including the identification information of the target person may be a sound emitted by a person different from the target person.

  Thereby, the identification information of the object person can be extracted from the voice emitted by the person different from the object person including the identification information of the object person. Therefore, it is possible to extract the identification information of the target person from the voice or the question voice which the person different from the target person speaks to the target person, the voice introducing the target person, and the like.

  In the information processing apparatus according to aspect 3 of the present invention, in the above aspect 2, a person different from the target person introduces the target person from the text generated by performing voice recognition of the voice data input from the voice input unit. It may further include an introduction sentence extraction unit 13 which extracts an introduction sentence to be output and outputs the introduction sentence to the identification information extraction unit.

  According to the above configuration, a person different from the target person extracts the identification information of the target person from the introduction sentence introducing the target person. Thereby, for example, an introduction sentence can be extracted from the voice when a person different from the target person introduces the target person to a third party or an information processing apparatus. In addition, since the introductory sentence in which a person different from the subject introduces the subject is a sentence having a sentence pattern in a predetermined format including the identification information of the subject, it is possible to selectively extract the introductory sentence from the voice. The identification information of the subject can be extracted efficiently.

  In the information processing apparatus according to aspect 4 of the present invention, in any of the above aspects 1 to 3, the identification information is a personal name indicating the target person, and the instruction generation unit is the identification information extraction unit The operation instruction may be generated including the person name extracted as identification information.

  According to the above configuration, the instruction including the person's name of the subject is generated. This makes it possible to clarify and output to whom the operation instruction is directed.

  The information processing apparatus according to aspect 5 of the present invention is the unique information extraction unit for extracting the unique information unique to the target person from the face image obtained by imaging the face of the target person in any one of the above aspects 1 to 4 The above target person further comprising: a characteristic data extraction unit 17), wherein the identification information of the target person who has performed the predetermined operation is associated with the unique information of the target person extracted by the unique information extraction unit Registration data relating to the above may be stored in the storage unit (storage unit 20, registration database 24).

  According to the above configuration, unique information specific to the subject is extracted from the face image obtained by imaging the face of the subject, and identification information of the subject and unique information are associated and stored. Generally, the face is different for each subject, so that unique information unique to each subject can be extracted from the captured face image. As a result, the face of the subject can be imaged without being confused with another person, so that registration data regarding the subject can be accurately managed.

  A control method of an information processing apparatus 1 according to a sixth aspect of the present invention is a control method of an information processing apparatus including an audio input unit 31 that receives an input of audio, and includes identification information of a target person from the audio input unit. A voice input step (S1) for accepting voice input, an identification information extraction step (S5) for extracting identification information of the subject from the voice data of the voice received in the voice input step; Operation instruction for generating a predetermined operation including the identification information extracted in the identification information extraction step, an instruction generation step (S6), and the operation instruction generated in the instruction generation step And an instruction presenting step (S8) presented to the subject. According to the above-mentioned composition, the same effect as the above-mentioned mode 1 is produced.

  The information processing apparatus according to each aspect of the present invention may be realized by a computer. In this case, the computer is operated as each unit (software element) included in the information processing apparatus, and the information processing apparatus is operated by the computer. A control program of an information processing apparatus to be realized and a computer readable recording medium recording the same also fall within the scope of the present invention.

  The present invention is not limited to the above-described embodiments, and various modifications can be made within the scope of the claims, and embodiments obtained by appropriately combining the technical means disclosed in the different embodiments. Is also included in the technical scope of the present invention. Furthermore, new technical features can be formed by combining the technical means disclosed in each embodiment.

  The present invention can be used for electronic devices, robots, etc. provided with a communication function with human beings.

1 information processing apparatus 13 introduction sentence extraction unit 14 person name extraction unit (identification information extraction unit)
15 Voice instruction generator (instruction generator)
17 Feature data extraction unit (specific information extraction unit)
18 Registered data management unit 20 Storage unit 24 Registration database (storage unit)
31 voice input unit 32 voice output unit (instruction presenting unit)
33 Imaging device

Claims (5)

An information processing apparatus comprising an audio input unit for receiving an input of audio, the information processing apparatus comprising:
An identification information extraction unit for extracting identification information of the target person from voice data of speech including identification information of the target person input from the voice input unit;
An instruction generation unit that generates an operation instruction for instructing the subject to perform a predetermined operation including the identification information extracted by the identification information extraction unit;
Providing an instruction presentation unit for presenting the operation instruction to the target person,
From the text generated by speech recognition of voice data input from the voice input unit, an introduction sentence in which a person different from the subject person introduces the target person is extracted, and the introduction sentence is output to the identification information extraction unit An information processing apparatus further comprising an extraction unit. The identification information is a personal name indicating the target person,
The information processing apparatus according to claim 1, wherein the instruction generation unit generates the operation instruction including the personal name extracted by the identification information extraction unit as the identification information. It further comprises a unique information extraction unit for extracting unique information unique to the subject from the face image obtained by imaging the face of the subject,
Storing in the storage unit registration data regarding the target person in which the identification information of the target person who has performed the predetermined operation and the unique information of the target person extracted by the unique information extraction unit are associated with each other The information processing apparatus according to claim 1 or 2, characterized in that   A control program for causing a computer to function as the information processing apparatus according to claim 1, wherein the control program causes the computer to function as the identification information extraction unit, the instruction generation unit, and the introduction sentence extraction unit. The computer-readable recording medium which recorded the control program of Claim 4 .

Images