JP5332602B2 - Service providing equipment - Google Patents

Description

  The present invention relates to a mobile service providing apparatus that transmits and receives information by voice.

  Recently, a bipedal robot that provides various services to humans has been developed. This biped robot transmits and receives information wirelessly with a server (see Non-Patent Document 1 and Non-Patent Document 2).

  In addition, an invention related to an interactive robot that performs a voice conversation with a human and provides a service according to the content of the conversation is disclosed (see Patent Document 1). In the present invention, robots cooperate with each other by transmitting and receiving information using a predetermined communication protocol to provide services to humans.

Further, an invention is disclosed in which identifiers and instruction information are superimposed on voices uttered by a robot, and information communication is performed between the robots by voice conversation (see Patent Document 2).
Robot Watch ASIMO, taking a step forward to the real world-Release of coordinated operation by multiple units, [online] [Search December 25, 2008], Internet <URL: http://robot.watch.impress.co.jp/cda /news/2007/12/12/793.html> Robot Watch ASIMO, taking a step forward to the real world-publicizing coordinated operation by multiple units System overview. Dynamically allocate functions (enlarged photo) [online] [searched December 25, 2008], Internet <URL: http: //robot.watch.impress.co.jp/cda/parts/image_for_link/23110- 793-10-1.html> International Publication No. 2005/086051 Pamphlet JP 2000-207170 A

  As described above, the robots described in Non-Patent Document 1 and Non-Patent Document 2 and the robot described in Patent Document 1 transmit and receive information by communication without having a voice conversation with a server or another robot. For this reason, there is a problem that a person who receives the service cannot understand what information is exchanged between the robots, and becomes anxious.

  Also, if multiple robots are all of the same model and the voices of the synthesized speech are the same, the synthesized speech produced by each robot cannot be distinguished from each other, so it is impossible to grasp which robot uttered, resulting in a hindrance to communication. There was a problem of coming.

  Further, when information is transmitted and received between only a plurality of robots as in Patent Document 2, the amount of information is small compared to the case where information is transmitted and received by wireless communication or the like. There was a problem that it could not be delivered.

  Accordingly, the present invention provides a service providing apparatus that can understand the content of information to be transmitted / received between a plurality of apparatuses even for a person receiving service provision, can identify even the same voice, and can transmit / receive all necessary information. The purpose is to provide.

  When the service providing apparatus according to the present invention picks up the sound including the identification information of the other apparatus by the sound collecting means when the sound is emitted to the other apparatus by the sound emitting means, the identification information Predetermined information is transmitted / received to / from other devices corresponding to the above by means of sound emission means and sound collection means, and auxiliary information is transmitted / received by communication means.

  As a result, the person who receives the service can grasp the contents of the information, and can identify which device has the voice even if the voices are the same. Further, even when not all information can be transmitted / received by voice alone, all necessary information can be transmitted / received.

  The service providing apparatus of the present invention directs the sound emitting means in the direction of arrival of the sound detected by the direction detecting means. As a result, the service providing apparatus can emit the sound toward the speaker, so that the operation is the same as when a person has a conversation, and the user does not feel uncomfortable.

  In addition, the service providing apparatus according to the present invention superimposes information by controlling a polarity of the pseudo noise signal by using a part of a sample signal for speech synthesis as a pseudo noise signal. Thereby, identification information of an individual can be transmitted and received without giving a sense of incongruity to a person.

  According to the present invention, since information is transmitted and received between a plurality of service providing apparatuses using voice on which individual identification information is superimposed, the content of the information can be grasped by the person receiving the service, and the same voice Even so, it is possible to identify which device is uttered. In addition, when the extraction unit extracts identification information from the voice of another device, information is transmitted / received to / from the other device, and information is transmitted / received to / from the device by the communication unit. Even in this case, all necessary information can be transmitted and received.

  Hereinafter, a service providing system including the service providing apparatus of the present invention will be described. In the following description, a robot that is an example of a service providing apparatus will be described as an example. FIG. 1 is a block diagram of a service providing system.

  In the present embodiment, the A / D converter and the D / A converter are omitted, and unless otherwise specified, all digital processing will be described.

  As shown in FIG. 1, the service providing system 1 has a configuration including robots 2 to 4 as an example. The robots 2 to 4 have a voice conversation with the user 5 to whom the service is provided, and provide various services according to the content of the conversation. Since the robots 2 to 4 have the same configuration, the configuration of the robot 2 will be described below as a representative example. It is assumed that the robot 3 exists in a range where the voice uttered by the robot 2 can reach and the robot 4 exists on the opposite side of the wall 6 where the voice uttered by the robot 2 does not reach.

  The robot 2 includes a sensor unit 11, a conversation engine 12, a speech unit 13, a storage unit 14, a reception unit 15, a communication unit 16, an operation unit 17, and a control unit 18.

  In the robot 2, the control unit 18 controls the conversation engine 12 to output a sound corresponding to the human movement detected by the sensor unit 11, the contents of the voice information recognized by the receiving unit 15 and the conversation engine 12, and the like. Output a signal.

  The conversation engine 12 outputs an utterance sentence (text information) created according to a control signal from the control unit 18 to the utterance unit 13.

  The utterance unit 13 synthesizes speech based on the text information input from the conversation engine 12, superimposes the individual ID (individual identification information) read from the storage unit 14 on the synthesized speech, and externally uses a speaker (not shown). Sounds out. In addition, when superimposing individual ID on a synthetic | combination voice, it is good to superimpose, after processing so that a sense of incongruity may not be given. For example, the individual ID may be converted into an acoustic watermark and superimposed (for example, superimposed on the ultrasonic band).

  Further, the utterance unit 13 superimposes predetermined information input from the control unit 18 on the synthesized speech and emits the sound to the outside through a speaker.

  The storage unit 14 stores an individual ID, information on other robots 3 and 4, and the like.

  In the robot 2, voices uttered by other robots 3, 4 (other devices) and users 5 (human beings who provide services) are picked up by a microphone (not shown) of the receiver 15 to be recognized. And outputs the result to the conversation engine 12. In addition, the receiver 15 extracts and decodes the individual ID superimposed on the audio signal, and if the individual ID can be extracted, outputs the result of decoding which robot the individual ID belongs to to the control unit 18. To do. Further, if the receiving unit 15 cannot extract the individual ID, the receiving unit 15 outputs the fact to the control unit 18. The robot 2 is provided with a plurality of microphones (at least two) in order to detect the direction of arrival of voice.

  The conversation engine 12 analyzes the input content and outputs the analysis result to the control unit 18.

  The control unit 18 performs processing according to the analysis result by the conversation engine 12. Further, processing according to predetermined information superimposed on the audio signal and processing according to information received via the communication unit 16 are performed. For example, when the user 5 utters “move to living room”, the conversation engine 12 performs analysis and performs a process of moving to the living room. Further, when a voice “move to living room” is issued from another device (for example, the robot 3), a process of moving to the living room is performed based on the information of the movement instruction superimposed on the voice. At this time, status information (remaining battery capacity or location) is transmitted / received as auxiliary information to / from other devices via the communication unit 16.

  That is, the control unit 18 can extract the individual ID from the voice signal when the voice including the identification information of the other robot is picked up when the voice of the content called to the other robot is emitted. In the case, the robot (robot 3) corresponding to the ID and the voice of the content corresponding to the analysis result of the conversation engine 12 are output to the conversation engine 12 to transmit / receive information (conversation), and the communication unit 16 Send and receive information. Alternatively, the control unit 18 controls the operation unit 17 according to the analysis result of the conversation engine 12 to perform a predetermined operation.

  In addition, when the control unit 18 does not pick up the voice including the identification information of the other robot when the voice of the content called to the other robot is emitted (the robot 3 has also moved to the opposite side of the wall 6). In other words, if the individual ID cannot be extracted from the audio signal, it is determined that there is no other robot in the range where the audio can reach, and the communication unit 16 exchanges information with other robots (robot 3 and robot 4). I do. Note that the communication method of the communication unit 16 may be wireless or wired.

  Further, when the control unit 18 does not emit the calling voice and picks up the voice that does not include the identification information of other robots, the control unit 18 determines that the voice is that of the user 5 and the conversation engine 12 The voice of the content corresponding to the analysis result is output to the conversation engine 12 to have a conversation with the user 5, or the operation unit 17 is controlled to perform a predetermined operation.

  In addition, since the microphone 15 is provided in the receiver 15 as described above, the controller 18 receives the voice (the direction in which the robot 3 or the user 5 is speaking) when the voice is picked up. ) Can be detected. In other words, the control unit 22 can detect the voice arrival direction by comparing the gains of the voice signals collected by the plurality of microphones. When detecting the voice arrival direction, the control unit 22 controls the operation unit 17 to direct the speaker of the utterance unit 13 (corresponding to the mouth of the robot 2) in the voice arrival direction. Thereby, since the robot 2 can emit a sound toward the speaker, the operation is the same as that when a human talks, and the user does not feel uncomfortable.

  Next, details of the utterance unit 12 and the reception unit 15 will be described. FIG. 2A is a block diagram showing details of the utterance unit. FIG. 2B is a block diagram showing details of the receiver.

  The utterance unit 12 includes a pseudo noise generation unit 21, a polarity control unit 22, a speech synthesis unit 23, an amplifier 24, and a speaker 25.

  The pseudo noise generation unit 21 generates a pseudo noise code string (PN code, pseudo noise signal waveform) having a high autocorrelation such as an M sequence or a Gold sequence.

  The polarity control unit 22 superimposes the individual ID on the pseudo noise by controlling the polarity of the pseudo noise code string generated by the pseudo noise generation unit 21 according to the individual ID read from the storage unit 14. That is, when the bit data of the individual ID read from the storage unit 14 is “1”, the pseudo noise generation unit 21 outputs the PN code with the same polarity, and when the bit data of the individual ID is “0”, the pseudo noise The generation unit 21 outputs the PN code with the polarity reversed (reverse phase).

  On the receiving side, by detecting the phase of the calculated correlation value, “1” and “0” of the superimposed bit data (individual ID) can be demodulated.

  The speech synthesizer 23 synthesizes speech from the pseudo noise output from the pseudo noise generator 21 and other waveforms (other waveform generators are not shown) according to the utterance text output from the conversation engine 12.

  The voice signal synthesized by the voice synthesis unit 23 is amplified by the amplifier 24 and emitted as voice. The emitted sound is collected by the microphone 31 of another robot and input to the receiver 15 shown in FIG.

  As shown in FIG. 2B, the receiver unit 15 includes a microphone 31, a voice recognition unit 32, a matched filter 33, and a demodulation unit 34.

  The voice recognition unit 32 recognizes the voice picked up by the microphone 31 and outputs text information to the conversation engine 12.

  The matched filter 33 is a correlation calculation unit that calculates the correlation between the input audio signal and the pseudo noise. The matched filter 33 is realized by an FIR filter, and a pseudo noise code string (PN code) generated by the pseudo noise generating unit 21 on the transmission side is set as a filter coefficient. Since the PN code has very high autocorrelation, the matched filter 33 outputs a correlation value peak (correlation value of a predetermined level or higher) when the input speech includes the PN code. The matched filter 33 outputs a positive correlation value peak if the phase is normal, and outputs a negative correlation value peak if the phase is inverted.

  The demodulator 34 demodulates data from the output value of the matched filter 33. That is, the demodulator 34 demodulates “1” as bit data when a positive correlation value peak is input from the matched filter 33, and as bit data when a negative correlation value peak is input from the matched filter 33. Demodulate "0". Note that the output period of the pseudo noise is determined in advance, and when the correlation value peak is input, the demodulator 34 continues to output the bit by the length of the output period of the pseudo noise. For example, if the period of the pseudo noise is 1023 samples, “1” is continuously output for 1023 samples when a positive correlation value peak is input.

  In this way, the demodulator 34 demodulates the individual ID and other information superimposed on the pseudo noise. The pseudo noise is a noise sound such as white noise itself in terms of frequency characteristics, and can be completely equivalent to a normal noise sound in terms of hearing.

  Next, the operation of the robot will be described based on a flowchart. FIG. 3 is a flowchart for explaining the operation of the robot.

  When the robot 2 picks up the sound with the microphone 31 (s1: Y), the receiving unit 15 performs voice recognition, confirmation of presence / absence of the individual ID, and decoding (s2). If the individual ID is not included in the voice (s3: N), the robot 2 determines that the call is from the user 5, and confirms the result of voice recognition (request contents of the user 5) (s4). When the request content of the user 5 needs to be shared by a plurality of robots or when the request content needs to be transmitted to another robot, the robot 2 needs to transmit information to another robot (s5). : Y), a voice call is made to surrounding robots (s6). On the other hand, when the request content of the user 5 does not need to convey information to other robots (s5: N), the robot 2 performs a conversation or an action according to the content (s7). When the conversation or operation ends, the process returns to step s1.

  When the robot 2 picks up the sound with the microphone 31 (s1: Y), the receiving unit 15 performs voice recognition, confirmation of presence / absence of the individual ID, and decoding (s2). Then, when the individual ID is included in the voice (s3: Y), the robot 2 determines that the reaction is from another robot, and transmits / receives information by voice and communication, or performs an operation according to the information. (S8). When the conversation or operation ends, the process returns to step s1.

  On the other hand, the robot 2 waits for a certain time until the sound is picked up (s1: N). If the sound is not picked up by the microphone 31 even after the predetermined time has passed (s1: N, s9: Y), It is determined that there is no other robot within the reach of the voice, and information is transmitted / received by communication of the other robot (s10). When the transmission / reception of information ends, the process returns to step s1.

  As described above, when it is necessary to transmit information to other robots, the robot of the present invention makes a call by voice, and when there is a response from other robots, sends and receives information by voice. Therefore, since the robots have a conversation with each other, people (users) around the robot can know what information is being exchanged between the robots, so that it is possible to avoid discomfort and anxiety.

  In addition, since the robots can transmit and receive information by voice and can transmit and receive information by communication, even when necessary information cannot be transmitted by voice alone, the information can be reliably transmitted by communication. In addition, by transmitting the same information as the information to be transmitted by voice, the information is transmitted twice, so even if the voice recognition is wrong, the robot can reliably operate based on the information without malfunctioning. Can be done. For example, when the user sends a voice “moving to living room” to the robot 2, the robot 2 calls the robot 3 when the remaining battery level is low and another robot 3 needs to be replaced. In addition, status information (auxiliary information) is transmitted and received via the communication unit 16. As a result, the robot 3 having a large remaining battery capacity performs a process of moving to the living room. Here, when the remaining battery level of the other robot 3 is less than that of the robot 2, the other robot 3 pronounces “Please move instead” and the robot 2 moves to the living room.

  In the above description, a robot picks up a voice and recognizes a robot uttered by detecting an ID from the voice, and then transmits / receives information by communication between the robots. It may be. For example, after confirming that communication is possible between robots, if it is necessary to transmit information, a voice is uttered and a call is made to surrounding robots. If detected, information is transmitted and received by voice and communication. If there is no response by voice, it is determined that there is no other robot within the reach of the voice, and information is transmitted / received to / from another robot. Even in this method, the surrounding people (users) can understand what information is being exchanged between the robots, so that it is not necessary to feel uncomfortable or uneasy.

  The service providing system 1 is not limited to three service providing apparatuses, and may be configured by a plurality of robots (service providing apparatuses).

  Although the robot has been described as an example of the service providing apparatus, the present invention is not limited to this, and the present invention can be widely applied to any information processing apparatus that performs communication by performing speech synthesis and speech recognition.

It is a block diagram of a service provision system. It is a block diagram of a speech part and a receiving part. It is a flowchart for demonstrating operation | movement of a robot.

Explanation of symbols

1 ... Service provision system 2-4 ... Robot

Claims (4)

A service providing device that transmits and receives information between a plurality of devices that perform communication by performing speech synthesis and speech recognition ,
Speech synthesis means for synthesizing speech;
Storage means for storing individual identification information;
Identification information superimposing means for superimposing the identification information on the voice synthesized by the voice synthesizing means;
A sound emitting means for emitting a sound on which the identification information is superimposed;
Sound collection means for collecting sound;
Extraction means for extracting identification information of another device from the sound collected by the sound collection means;
A communication means for performing information communication by connecting to another device;
When the identification information of another device is extracted by the extraction unit, the predetermined information is transmitted to the other device by superimposing the predetermined information on the voice synthesized by the voice synthesis unit, and the extracted identification information is added to the extracted identification information. connected via said communication means with corresponding other apparatuses, and control means for transmitting the auxiliary information to assist the predetermined information,
A service providing apparatus comprising:   The service providing apparatus according to claim 1, wherein the auxiliary information includes the same information as the predetermined information. Direction detecting means for detecting the direction of arrival of the sound collected by the sound collecting means;
The service providing apparatus according to claim 1, further comprising: an operation unit that directs the sound emitting unit toward an arrival direction of the sound detected by the direction detecting unit. The speech synthesis means is a means for synthesizing speech from a plurality of sample signals,
The identification information superimposition means, a portion of the sample signal as a pseudo-noise signal, the service providing apparatus according to any one of claims 1 to 3 superimposes the information by controlling the polarity of the pseudo noise signal .

Images