JP2002120177A - Robot control device, robot control method and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform suitable naming. SOLUTION: The voice input after a robot takes an action for expressing 'Please give a name' is recognized as a name, whereby a name is suitably given to the robot.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a robot control device, a robot control method, and a recording medium, and more particularly to a robot control device suitable for use in a robot acting on the basis of a result of voice recognition by a voice recognition device, for example. The present invention relates to a robot control method and a recording medium.

[0002]

2. Description of the Related Art In recent years, for example, as a toy or the like, a robot that performs voice recognition of a voice uttered by a user and performs a certain gesture based on the voice recognition result (in this specification, a stuffed toy is used). ) Have been commercialized.

[0003]

By the way, by giving a name to such a robot and calling the name,
A predetermined operation can also be performed.

[0004] However, when giving a name to a robot by causing a voice representing a name to be speech-recognized, the robot may erroneously recognize, for example, surrounding sounds produced before or after the user produces a voice representing the name. Sometimes it was recognized as a name.

If a word that is acoustically similar to a word registered in advance for voice recognition is registered as a name, even if the name is called, the robot will recognize the voice,
In some cases, the similar words were mistakenly recognized. Also, when the similar word is spoken, the robot may erroneously recognize that the name has been called.

That is, when giving a name to a robot by performing voice recognition as described above, there has been a problem that it is not possible to appropriately assign a name or to accurately recognize a name.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and has been made in view of the above circumstances. In the case where a name is given by voice recognition, the name can be appropriately given or the name can be accurately recognized. It is.

[0008]

SUMMARY OF THE INVENTION A robot control device according to the present invention comprises: a behavior control means for controlling a behavior of a robot so that the behavior of the robot indicates that a name is desired; The present invention is characterized in that it comprises a detecting means for detecting an optimal phoneme string from a voice inputted after the action is controlled, and a registering means for registering the optimal phoneme string as a name.

[0009] A transition means for sequentially transitioning the growth state of the robot to a predetermined state is further provided, and the behavior control means changes the name when the growth state of the robot is transitioned to the predetermined state by the transition means. It is possible to control the behavior of the robot so that the robot performs an action indicating that the robot wants to attach.

An acquisition means for acquiring a command input by the user for instructing the start of the name registration process is further provided, and the behavior control means is required to give a name when the acquisition means acquires the command. Can be controlled so that the robot takes the action representing

A first storage means for storing an acoustic model representing an acoustic feature, and a second storage means for storing a word dictionary in which words for speech recognition are registered, wherein the detection means are provided. The word model having the highest score at which the feature parameter is observed is detected as the optimal phoneme sequence among the word models formed of the acoustic models connected corresponding to the speech feature parameters. Among the word models consisting of acoustic models connected corresponding to the phonemes of the already registered words, a word model having the highest score in which the characteristic parameter is observed is detected. If the difference between the score of the phoneme string and the score is larger than a predetermined threshold, the optimum phoneme string can be registered as a name.

According to the robot control method of the present invention, the action of the robot is controlled by the action control step of controlling the action of the robot so that the action of the robot indicates that the name is desired. It is characterized by including a detecting step of detecting an optimal phoneme string from a voice inputted later and a registration step of registering the optimal phoneme string as a name.

[0013] The program of the recording medium of the present invention has a behavior control means for controlling the behavior of the robot and an action control means for controlling the behavior of the robot so that the behavior of the robot indicates that a name is desired. And a registration unit for registering the optimal phoneme string as a name from the speech input after the input.

In the robot control apparatus and method of the present invention, and the program of the recording medium, the behavior of the robot is controlled so that the robot performs an action indicating that a name is desired to be given, and the behavior of the robot is controlled. The optimal phoneme sequence is detected from the voice input later, and the optimal phoneme sequence is registered as a name.

[0015]

FIG. 1 shows an example of the appearance of a robot according to an embodiment of the present invention, and FIG.
An example of the electrical configuration is shown.

In this embodiment, the robot is, for example,
It has the shape of a four-legged animal such as a dog, and has leg units 3A,
3B, 3C, and 3D are connected, and a head unit 4 and a tail unit 5 are connected to a front end and a rear end of the body unit 2, respectively.

The tail unit 5 is drawn out from a base 5B provided on the upper surface of the body unit 2 so as to bend or swing with two degrees of freedom.

The body unit 2 contains a controller 10 for controlling the entire robot, a battery 11 as a power source of the robot, and an internal sensor unit 14 including a battery sensor 12 and a heat sensor 13.

The head unit 4 includes a microphone (microphone) 15 corresponding to “ears” and a CCD corresponding to “eyes”.
(Charge Coupled Device) A camera 16, a touch sensor 17 corresponding to "tactile sensation", and a speaker 18 corresponding to "mouth" are respectively provided at predetermined positions.
A D (Light Emitting Diode) 19 is provided at the position of the “eye”. Also, a lower jaw 4A corresponding to the lower jaw of the mouth is movably attached to the head unit 4 with one degree of freedom, and the opening and closing operation of the robot's mouth is realized by moving the lower jaw 4A. It has become.

The joint portions of the leg units 3A to 3D, the connecting portions of the leg units 3A to 3D and the body unit 2, the connecting portions of the head unit 4 and the body unit 2, the head unit 4 and the lower jaw linking moiety parts 4A, and the like in the connecting portion of the tail unit 5 and the body unit 2, as shown in FIG. 2, each actuator 3AA ₁ to 3AA _K, 3BA ₁ to 3BA _K, 3CA
₁ to 3CA _K, 3DA ₁ to 3DA _K, 4A ₁ to 4A _L,
5A ₁ and 5A ₂ are disposed.

The microphone 15 in the head unit 4 collects surrounding sounds (sounds) including utterances from the user, and sends out the obtained sound signals to the controller 10. CCD
The camera 16 captures an image of the surroundings, and sends the obtained image signal to the controller 10.

The touch sensor 17 is provided, for example, above the head unit 4 and “strokes” from the user.
It detects the pressure received by a physical action such as tapping or tapping, and sends the detection result to the controller 10 as a pressure detection signal.

The battery sensor 12 in the body unit 2 detects the remaining amount of the battery 11 and sends the detection result to the controller 10 as a battery remaining amount detection signal. The heat sensor 13 detects heat inside the robot,
The detection result is sent to the controller 10 as a heat detection signal.

The controller 10 has a CPU (Central Pro
(Processing Unit) 10A, a memory 10B, and the like. The CPU 10A performs various processes by executing a control program stored in the memory 10B.

That is, the controller 10 controls the microphone 15
And, based on sound signals, image signals, pressure detection signals, remaining battery level detection signals, and heat detection signals provided from the CCD camera 16, the touch sensor 17, the battery sensor 12, and the heat sensor 13, the surrounding conditions and the user Is determined, and whether or not there is a request from the user.

Furthermore, the controller 10, based on the determination results and the like, to determine the subsequent actions, based on the determination result, the actuators 3AA ₁ to 3AA _K, 3BA ₁
To 3BA _K, 3CA ₁ to 3CA _K, 3DA ₁ to 3DA
_K, 4A ₁ to 4A _L, 5A _1, 5A to drive the necessary of the _two. Thereby, the head unit 4 is swung up, down, left and right, and the lower jaw 4A is opened and closed. Further, the tail unit 5 can be moved, and the leg units 3A to 3D are driven to perform actions such as walking the robot.

Further, the controller 10 generates a synthesized sound or an echo-back sound as described later, as necessary, and supplies it to the speaker 18 for output, or is provided at the position of the "eye" of the robot. The LED 19 is turned on, turned off, or blinked.

As described above, the robot autonomously behaves based on the surrounding situation and the like.

FIG. 3 shows an example of a functional configuration of the controller 10 shown in FIG. Note that the functional configuration illustrated in FIG. 3 is realized by the CPU 10A executing a control program stored in the memory 10B.

The sensor input processing unit 50 includes a microphone 15,
Based on audio signals, image signals, pressure detection signals, and the like provided from the CCD camera 16, the touch sensor 17, and the like, a specific external state, a specific action from the user, an instruction from the user, and the like are recognized, and the recognition result is obtained. Is notified to the model storage unit 51 and the action determination mechanism unit 52.

That is, the sensor input processing unit 50 has a voice recognition unit 50A, and the voice recognition unit 50A
Speech recognition is performed on the speech signal given from 5. Then, the voice recognition unit 50A uses, as the state recognition information, commands such as “walk”, “down”, “chase the ball”, etc., and registered names and the like as described later as the voice recognition result. It notifies the model storage unit 51 and the action determination mechanism unit 52.

The sensor input processing section 50 has an image recognizing section 50B. The image recognizing section 50B performs an image recognizing process using an image signal supplied from the CCD camera 16. When the image recognition unit 50B detects, for example, a “red round object” or a “plane that is perpendicular to the ground and equal to or more than a predetermined height” as a result of the processing,
Image recognition results such as “there is a ball” and “there is a wall” are notified to the model storage unit 51 and the action determination mechanism unit 52 as state recognition information.

Further, the sensor input processing section 50 has a pressure processing section 50C, and the pressure processing section 50C processes a pressure detection signal given from the touch sensor 17. Then, when the pressure processing unit 50C detects a pressure that is equal to or more than a predetermined threshold value and for a short time as a result of the processing, the pressure processing unit 50C recognizes that “the head has been touched” and uses the recognition result as state recognition information. Model storage unit 51 and action determination mechanism unit 52
Notify.

The model storage unit 51 stores and manages an emotion model, an instinct model, and a growth model expressing the emotion, instinct, and growth state of the robot.

Here, the emotion model expresses emotion states (degrees) such as, for example, "joy,""sadness,""anger," and "fun," by values in a predetermined range. The value is changed based on the state recognition information from the input processing unit 50, the passage of time, and the like.

The instinct model is a state (degree) of desire by instinct such as “appetite”, “sleep desire”, and “exercise desire”.
Are respectively represented by values within a predetermined range, and the values are changed based on state recognition information from the sensor input processing unit 50, elapsed time, and the like.

The growth model is composed of, for example, an automaton as shown in FIG. In this automaton, the growth state is represented by nodes (states) NODE0 to NODEG, and the growth, that is, the transition of the growth state is from a node NODEg corresponding to a certain growth state to a node NODEg + 1 corresponding to the next growth state. ARCg + 1 representing the transition of
(G = 0, 1,..., G-1).

Here, in the present embodiment, the state of growth transitions from the left node to the right node in FIG. Therefore, in FIG.
For example, the leftmost node NODE0 represents the state of a "newborn" just born, and the second node NO.
DE1 represents the state of "child", and the third node NODE2 from the left represents the state of "child". Hereinafter, similarly, the rightward node represents a more mature state, and the rightmost node NODEG represents an "elderly" state.

An arc ARCg + 1 representing a transition from a certain node NODEg to the node NODEg + 1 on the right side is set with a condition (input) Ptg + 1 for causing the transition. Transition (growth) is determined based on this condition. That is, in the arc ARCg + 1,
CCD camera 16 required for the transition to occur
Ptg + 1 for the output from the microphone 9, the touch sensor 17, the passage of time, and the like are set.
When tg + 1 is satisfied, a transition from the node NODEg to the node NODEg + 1 on the right occurs, and the robot grows.

The model storage unit 51 sends the emotion, instinct, and growth state represented by the values of the emotion model, instinct model, and growth model as described above to the behavior determination mechanism unit 52 as state information.

The model storage unit 51 is supplied with state recognition information from the sensor input processing unit 50, and the current or past behavior of the robot, specifically, for example, “ Behavior information indicating the content of the action such as "walking for a long time" is supplied. Even if the same state recognition information is given, different state information ( (Emotion, instinct, and growth state) represented by the values of the emotion model, the instinct model, and the growth model.

The action determining mechanism 52 determines the next action based on the state recognition information from the sensor input processing section 50, the state information from the model storage section 51, the passage of time, and the like. The content is sent to the posture transition mechanism 53 as action command information.

That is, the action determining mechanism 52 manages a finite state automaton in which actions that can be taken by the robot correspond to states, as an action model that defines the actions of the robot. State in the finite state automaton as the sensor input processing unit 5
Transition based on state recognition information from 0, the value of the emotion model, instinct model, or growth model in the model storage unit 51, the passage of time, and the like, and the action corresponding to the state after the transition as the action to be taken next decide.

Here, upon detecting that a predetermined trigger has occurred, the action determining mechanism 52 changes the state. That is, for example, when the time during which the action corresponding to the current state is being executed reaches a predetermined time, or when specific state recognition information is received, the action determining mechanism unit 52 is supplied from the model storage unit 51. The state is changed when the value of the emotion, instinct, or growth state indicated by the state information is equal to or less than a predetermined threshold.

As described above, the behavior determining mechanism 52 stores not only the state recognition information from the sensor input processing unit 50 but also the values of the emotion model, the instinct model, the growth model, and the like in the model storage unit 51. Based on the state transition based on the behavior model, the destination of the state transition differs depending on the emotion model, the instinct model, and the value of the growth model (state information) even if the same state recognition information is input. Become.

As a result, for example, when the state information indicates “not angry” and “not hungry”, the action determining mechanism 52 sets the state recognition information to “eye”. When the palm has been presented before, the action command information for taking the action of "hand" is generated in accordance with the palm being presented in front of the eyes, It is sent to the posture transition mechanism 53.

Further, for example, when the state information indicates that “the person is not angry” and “is hungry”, the action determining mechanism 52 determines that the state recognition information indicates “the front of the eyes”. When the palm is displayed, the action command information for performing an action such as "palm licking the palm" is generated in response to the palm being displayed in front of the eyes. This is sent to the posture transition mechanism 53.

Further, for example, when the state information indicates “angry”, the action determining mechanism 52 determines that the state recognition information indicates “the palm is put in front of the eyes”. When it indicates, even if the status information indicates that "stomach is hungry", or indicates that "stomach is not hungry", even if the state information indicates "being hungry", an action such as "turns to the side with a little bit" The action command information for performing the action is generated and sent to the posture transition mechanism 53.

The behavior determining mechanism 52 has an emotion, an instinct, and the like indicated by the state information supplied from the model storage 51.
Based on the state of growth, as a parameter of the action corresponding to the state of the transition destination, for example, the speed of walking, the magnitude and speed of the movement when moving the limbs can be determined, in this case, Is transmitted to the posture transition mechanism 53.

In addition, as described above, the action determining mechanism 52 generates action command information for causing the robot to speak, in addition to action command information for operating the robot's head and limbs. The action command information that causes the robot to speak is
The voice command is supplied to the voice synthesis unit 55, and the action command information supplied to the voice synthesis unit 55 includes the voice synthesis unit 5.
5 includes a text corresponding to the synthesized sound to be generated. Then, upon receiving the action command information from the action determination mechanism section 52, the speech synthesis section 55 generates a synthesized sound based on the text included in the action command information, and outputs the synthesized sound.
7 to the speaker 18 for output. As a result, for example, the cry of the robot, various requests to the user such as “hungry”, a response to the user's call such as “what?”, And other voice output are performed from the speaker 18. .

The posture transition mechanism unit 53 includes the action determination mechanism unit 5
Based on the action command information supplied from 2, posture change information for changing the posture of the robot from the current posture to the next posture is generated and transmitted to the control mechanism unit 54.

Here, the posture that can be changed next from the current posture is, for example, the physical shape of the robot such as the shape and weight of the body, hands and feet, the connection state of each part, the direction in which the joint is bent, and the like. It is determined by the mechanism of the actuator 3AA ₁ to 5A ₁ and 5A _2, such as angle.

As the next posture, there are a posture that can directly transition from the current posture and a posture that cannot directly transition. For example, a four-legged robot can make a direct transition from lying down with its limbs throwing down to lying down, but not directly into a standing state. It is necessary to perform a two-stage operation of pulling down to a prone position and then standing up. There are also postures that cannot be safely executed. For example, a four-legged robot easily falls down when trying to banzai with both front legs raised from its standing posture.

For this reason, the posture transition mechanism unit 53 pre-registers a posture to which a direct transition is possible, and when the action command information supplied from the behavior determination mechanism unit 52 indicates a posture to which a direct transition is possible, The action command information is sent to the control mechanism unit 54 as posture change information as it is. on the other hand,
When the action command information indicates a posture that cannot directly make a transition, the posture transition mechanism unit 53 generates posture transition information that makes a transition to a target posture after temporarily transiting to another possible posture. To the control mechanism 54. As a result, it is possible to avoid a situation in which the robot forcibly executes an untransitionable posture or a situation in which the robot falls.

In accordance with the posture transition information from the posture transition mechanism 53, the control mechanism 54
₁ generates a control signal for driving the 5A ₁ and 5A _2, which is sent to the actuator 3AA ₁ to 5A ₁ and 5A _2. Thereby, the actuator 3AA ₁
To 5A ₁ and 5A ₂ is driven in accordance with the control signals, the robot causes the autonomous motions.

The echo back unit 56 monitors a voice signal provided from the microphone 15 and recognized by the voice recognition unit 50A, and reproduces the voice signal (hereinafter referred to as echo back voice as appropriate). ) Is generated and output. This echo back sound is supplied to the speaker 18 via the output control unit 57 and output.

The output control section 57 is supplied with the digital data of the synthesized sound from the voice synthesis section 55 and the digital data of the echo-back sound from the echo-back section 56. Is converted into an analog audio signal by D / A and supplied to the speaker 18 for output. When the output of the synthesized sound from the voice synthesizer 55 and the echo-back sound from the echo-back unit 56 to the speaker 18 conflicts, the output control unit 57 adjusts the conflict. That is, the output of the echo-back sound from the echo-back unit 56 is performed independently of the output of the synthesized sound performed by the voice synthesizing unit 55 under the control of the action determination mechanism unit 52. The output and the output of the synthesized sound may conflict with each other. Therefore, the output control unit 57 arbitrates the conflict.

The name registration unit 58 is provided with a sensor input processing unit 50
(Speech recognition unit 50A, pressure processing unit 50C), model storage unit 51, action determination mechanism unit 52, echo back unit 56, and LED 19 are controlled to execute a name registration process described later.

Next, FIG. 5 shows an example of the configuration of the voice recognition section 50A of FIG.

The audio signal from the microphone 15 is AD (Analo
g Digital) converter 21. The AD converter 21 samples and quantizes an audio signal, which is an analog signal from the microphone 15, and A / D converts the audio signal into digital audio data. This audio data is supplied to the feature extraction unit 22 and the audio section detection unit 27.

The feature extracting unit 22 converts, for example, MF
A CC (Mel Frequency Cepstrum Coefficient) analysis is performed, and the analysis result is used as a feature parameter (feature vector).
Is output to the matching unit 23. The feature extraction unit 22 can also extract, for example, a linear prediction coefficient, a cepstrum coefficient, a line spectrum pair, power (output of a filter bank) for each predetermined frequency band, and the like as feature parameters.

The matching section 23 uses the feature parameters from the feature extraction section 22 to refer to the acoustic model storage section 24, the dictionary storage section 25, and the grammar storage section 26 as necessary, and to be input to the microphone 15. Voice (input voice)
Is recognized based on, for example, a continuous distribution HMM (Hidden Markov Model) method.

That is, the acoustic model storage unit 24 stores acoustic models representing acoustic features such as individual phonemes and syllables in the language of the speech to be recognized. Here, since speech recognition is performed based on the continuous distribution HMM method, HMM (Hidden Markov Model) is used as an acoustic model.

As shown in FIG. 6, the dictionary storage unit 25 stores a word dictionary in which information on pronunciation (phonological information) is described for each word to be recognized. The grammar storage unit 26 stores grammar rules describing how each word registered in the word dictionary of the dictionary storage unit 25 is linked (connected). Here, as the grammar rule, for example, a rule based on a context-free grammar (CFG), a statistical word chain probability (N-gram), or the like can be used.

The matching section 23 refers to the word dictionary in the dictionary storage section 25 and connects the acoustic models stored in the acoustic model storage section 24 to form a word acoustic model (word model). .

Further, the matching unit 23 connects several word models by referring to the grammar rules stored in the grammar storage unit 26, and uses the thus connected word models to generate feature parameters. , The speech input to the microphone 15 is recognized by the continuous distribution HMM method. That is, the matching unit 23 detects the sequence of the word model having the highest score (likelihood) at which the time-series feature parameters output by the feature extraction unit 22 are observed, and determines the word sequence corresponding to the word model sequence. The phoneme information (reading) is output as a speech recognition result.

The microphone 15 output as described above
The recognition result of the voice input to the
It is output to the model storage unit 51 and the action determination mechanism unit 52.

The voice section detecting section 27 calculates, for example, the power of the voice data from the AD converting section 21 for each frame similar to that in which the feature extracting section 22 performs the MFCC analysis. Further, the voice section detection unit 27 compares the power of each frame with a predetermined threshold, and detects a section including frames having power equal to or higher than the threshold as a voice section in which the user's voice is input. . Then, the voice section detecting section 27 supplies the detected voice section to the feature extracting section 22 and the matching section 23, and the feature extracting section 22 and the matching section 23 perform processing only on the voice section.

FIG. 7 shows an example of the configuration of the echo back unit 56 of FIG.

The audio signal from the microphone 15 is supplied to the AD converter 41. The A / D converter 41 samples and quantizes an audio signal, which is an analog signal from the microphone 15, and A / D-converts the audio signal into digital audio data. The voice data is supplied to the prosody analysis unit 42 and the voice section detection unit 46.

The prosody analysis unit 42 extracts the prosody information of the voice data such as the pitch frequency and the power by acoustically analyzing the voice data input thereto for each appropriate frame. This prosody information is stored in the sound generation unit 4
3 is supplied.

The sound generation unit 43 generates an echo-back sound whose prosody is controlled based on the prosody information from the prosody analysis unit 42.

That is, the sound generation unit 43 outputs a phoneme-free speech (hereinafter, appropriately referred to as a silent phoneme speech) having the same prosody as the prosody information from the prosody analysis unit 42, for example, as a sign
The signal is generated by superimposing (sin) waves and supplied to the output unit 44 as echo back sound.

Note that methods for generating voice data from, for example, pitch frequency and power as prosodic information are described in, for example, Suzuki, Ishii, Takeuchi, "Echoing imitation by non-segmented sounds and its psychological effect", IPSJ Transactions, vo
l.41, No.5, pp1328-1337, May, 2000, and JP-A-2000-181896
The details are described in Japanese Patent Publication No.

The output unit 44 stores the echo back sound data from the sound generation unit 43 in the memory 45 and outputs the data to the output control unit 57 (FIG. 3).

The voice section detection section 46 detects the voice section by performing the same processing as that in the voice section detection section 27 of FIG. It is supplied to the sound generation unit 43. As a result, the prosody analysis unit 42 and the sound generation unit 43 perform processing only on the voice section.

The AD converter 41 or the voice section detector 46 shown in FIG. 7 can be used as the AD converter 21 or the voice section detector 27 shown in FIG.

The echo back unit 5 configured as described above
In step 6, for example, an echo back process is performed according to the flowchart of FIG.

That is, first, in step S 11, the voice section detection unit 46 determines whether or not the voice section is a voice section based on the output of the AD conversion unit 41. Finished and step S again
The echo back process from step 11 is restarted.

If it is determined in step S11 that the voice section is a voice section, the process proceeds to step S12, where the prosody analysis section 42 outputs the output of the AD conversion section 41, that is, the microphone 1
The prosody information is obtained by acoustically analyzing the user's voice input to 5 and supplied to the sound generation unit 43.

In step S 13, the sound generation unit 43 generates a silent sound having the same prosody as the prosody information from the prosody analysis unit 42, and supplies it to the output unit 44 as an echo-back sound.

The output unit 44 determines in step S14
The data of the echo-back sound from the sound generation unit 43 is stored in the memory 45, and the process proceeds to step S15, where the echo-back sound is output to the output control unit 57 (FIG. 3), and the process ends.

Thus, the echo back sound is supplied to the speaker 18 via the output control unit 57 and output.

Therefore, in this case, the speaker 18 outputs
From the voice uttered by the user, the one that lost the phoneme,
Output as echo back sound.

The echo back voice is transmitted to the voice recognition unit 50.
A is to repeat the voice of the user whose voice is to be recognized in A. As a result of outputting such an echo-back voice, the user listens to the echo-back voice, and the robot hears his / her own voice. Can be recognized. Therefore, when the robot does not take any action as a response to the voice from the user (including both the case where the voice of the user is correctly recognized and the case where the voice of the user is incorrectly recognized in the voice recognition unit 50A) ), But in the user,
It is possible to prevent the robot from misunderstanding that the robot is out of order.

Further, since the echo-back sound is not the sound itself uttered by the user but the phonology of the sound, the robot understands the user's voice and repeats with his own voice. Sounds as if you are. Therefore, in the robot, the user's voice is not simply recorded and played, but can be given to the user an impression as if he / she understands it.

Here, in the sound generation section 43,
The echo back sound is generated by superimposing the sine wave.In addition, for the echo back sound, for example, a complicated waveform that can be a voice of a robot is prepared and the waveforms are connected. Can be generated by Further, as the echo back voice, for example, it is possible to recognize a phoneme constituting a user's voice and generate a voice having a phoneme constituted by the phoneme sequence. Also, the echo-back sound can be generated, for example, by obtaining a cepstrum coefficient for the user's voice and using a digital filter that uses the cepstrum coefficient as a tap coefficient.

However, if the echo-back sound is too similar to the user's voice, the robot may give the user a so-called impressive impression as if the user's voice were simply recorded and reproduced. It is desirable that the echo-back sound should not be very similar to the user's sound.

In the above case, the sound generation unit 43 generates an echo-back sound having the same prosody as that of the user's voice. It is also possible to generate an echo back voice having a prosody obtained by adding some processing to the prosody.

Next, the procedure of the name registration process performed by the name registration unit 58 of FIG. 3 will be described with reference to the flowchart of FIG. The name registration process is started when it is determined in step S21 that the robot has grown to a state requiring a name, and the fact is notified to the name registration unit 58 in step S22.

More specifically, the action determining mechanism 52 determines whether the growth model as state information sent from the model storage 51 indicates a growth state requiring a name. Determine whether or not.

For example, humans and dogs can recognize their names once they have grown to some extent. Therefore, in this example, “infant” (FIG. 4) in the growth model is
When the growth model from the model storage unit 51 indicates “infant”, the robot has grown to a state that requires a name. And

If it is determined in step S21 that the robot has not grown to a state that requires a name (if the growth model represents a state where the robot has not grown from "infant"), an action is determined. The mechanism unit 52 ends the process, and restarts the process from step S21 again.

If it is determined in step S21 that the robot has grown to a state that requires a name (if the growth model represents “infant”), the process proceeds to step S22, where the action determining mechanism unit 52 To that effect, the name registration unit 5
Notify 8.

In this way, when the name registration unit 58 is notified that the robot has grown to a state requiring a name, in step S23, the name registration unit 58
By controlling the voice recognition unit 50A of the sensor input processing unit 50,
The voice recognition of the voice signal given from the microphone 15 is stopped. Thereby, step S29, which will be described later,
Until speech recognition is resumed at 38, no speech recognition is performed. At this time, the name registration unit 58 controls the echo back unit 56 to prohibit the execution of the echo back process (FIG. 8).

In step S24, the name registration unit 58
Controls the action determination mechanism 52 to cause the attitude transition mechanism 53 to output action command information indicating "I want you to touch your head." Thereby, based on the action command information from the action determination mechanism section 52, the attitude change mechanism section 53 changes the attitude of the robot to each of the actions in the action indicating “I want you to touch my head”. Generate information,
It is sent to the control mechanism 54. Control mechanism unit 54 in accordance with the posture transition information from the attitude transition mechanism part 53 generates control signals for driving the actuators 3AA ₁ to 5A ₁ and 5A _2, which, actuators 3AA ₁ to 5A ₁ and 5A ₂ To send to.

Actuators 3AA _{1 to} 5A ₁ and 5
A ₂ is driven according to the control signal, and the robot takes an action indicating “I want you to touch the head” as shown in FIG. 10 (the robot hits the head several times with its own hand). ).

Next, in step S25, the name registration unit 58 resets and starts the timer T built in the controller 10.

In step S26, the name registration unit 58
Communicates with the pressure processing unit 50C of the sensor input processing unit 50 to determine whether or not the pressure processing unit 50C recognizes that “the head has been touched”, that is, whether or not the user has touched the head of the robot. Is determined.

If it is determined in step S26 that the user has not touched the head of the robot, the flow advances to step S27, and the name registration unit 58 determines whether the value of the timer T started in step S25 is 10 or more. Is determined (whether or not 10 seconds have elapsed). If it is determined that 10 seconds have not elapsed, the process returns to step S26, and the subsequent processing is executed.

On the other hand, if it is determined in step S26 that the head has been touched, the flow advances to step S28, where the name registration unit 58 controls the action determination mechanism unit 52 to indicate that "I want a name to be given." The action command information to be represented is output to the posture transition mechanism 53. Thereby, the posture transition mechanism 53
Generates posture transition information for transitioning the posture of the robot to each posture in the behavior indicating “I want to give a name” based on the behavior command information from the behavior determination mechanism unit 52, 54. Control mechanism 5
4, in accordance with the posture transition information from the attitude transition mechanism part 53 generates control signals for driving the actuators 3AA ₁ to 5A ₁ and 5A _2, which is sent to the actuator 3AA ₁ to 5A ₁ and 5A ₂ .

Actuators 3AA _{1 to} 5A ₁ and 5
A ₂ is driven in accordance with the control signals, the robot, for example, as shown in FIG. 11, take action indicating that "I want to name" (robot, their ears, the above-friendly, it Shake left and right several times).

Next, in step S29, the name registration section 58 controls the speech recognition section 50A of the sensor input processing section 50 to restart the speech recognition stopped in step S23. At this time, the name registration unit 58 controls the echo back unit 56 to permit execution of the processing from step S11 to step S14 in the processing shown in FIG. As a result, the data of the echo-back sound generated by the input sound is stored in the memory 45 (although the processing of steps S11 to S14 is executed), and the echo-back sound is output from the speaker 18. Is not performed (the process of step S15 is not performed).

In step S30, name registration unit 58
Resets and starts the timer T.

Next, in step S31, the name registration unit 58 communicates with the voice recognition unit 50A of the sensor input processing unit 50 to determine whether a voice signal has been input to the voice recognition unit 50A.

If it is determined in step S31 that a voice signal is not input, the process proceeds to step S32, where the name registration unit 58 determines whether the value of the timer T started in step S30 is 10 (whether or not 10 seconds have elapsed). Or judge)
If it is determined that 10 seconds have not elapsed, step S31
To execute the subsequent processing.

If it is determined in step S31 that a voice signal has been input, the flow advances to step S33, in which the name registration unit 58 communicates with the voice recognition unit 50A, and proceeds to step S31.
It is confirmed whether or not the voice (name) input in is appropriate as the robot name. Details of the processing here are shown in the flowchart of FIG.

In step S51, the voice recognition unit 50
The A matching unit 23 (FIG. 5) converts the feature parameters supplied from the feature extraction unit 22 (from the audio data obtained by sampling the audio signal input in step S31 by the AD conversion unit 21) into the feature extraction unit 22. The acoustic model of the acoustic model storage unit 24 is connected to the acoustic model of the word (word model) in accordance with
Then, the matching unit 23 detects a word model (optimal phoneme sequence) having the highest score (likelihood) at which the feature parameter is observed, from the generated acoustic models.

Next, in step S52, the matching unit 23 refers to the word dictionary in the dictionary storage unit 25, connects the acoustic models stored in the acoustic model storage unit 24, and Model). Then, the matching unit 23 detects, from the acoustic models generated here, a word model (optimal registered phoneme sequence) having the highest score at which the characteristic parameter is observed.

In step S53, matching unit 2
3 calculates the difference between the score of the optimal phoneme string detected in step S51 and the score of the optimal registered phoneme string detected in step S53 (in this example, the score of the optimal phoneme string is
The score of the optimally registered phoneme sequence is subtracted), and in step S54, it is determined whether the calculation result (subtraction result) is greater than a predetermined threshold.
Go to 55.

At step S55, matching unit 2
No. 3 notifies the name registration unit 58 that the optimal phoneme string detected in step S51 is appropriate as a name. In this case, a word that is acoustically similar to the name represented by the input voice is not registered in the word dictionary of the dictionary storage unit 25 of the voice recognition unit 50A. Appropriate.

On the other hand, in step S54, step S53
If it is determined that the calculation result (subtraction result) is equal to or smaller than the predetermined threshold, the process proceeds to step S56, and the matching unit 23 determines that the optimal phoneme string detected in step S51 is not appropriate as a name (not Is appropriate), and notifies the name registration unit 58. In this case, a word acoustically similar to the name represented by the input voice is recognized by the voice recognition unit 50A.
Is registered in the word dictionary of the dictionary storage unit 25,
The detected optimal phoneme sequence is not appropriate as a name.

After the processing in step S55 or S56, the confirmation processing ends, and the flow advances to step S34 in FIG.

In step S34, name registration unit 58
Is based on the confirmation result in step S33 (in step S55 or step S56 in FIG.
Based on the notification content from 0A), it is determined whether the word represented by the voice input in step S31 (the optimal phoneme string detected in step S51) can be registered as a name.

If the name is notified in step S54 of FIG. 12 that the name is appropriate, the name registration unit 58 determines in step S34 that the name can be registered, and the process proceeds to step S35, where the speech recognition unit 50A Under the control, the optimal phoneme string detected in step S51 is registered as a name in the word dictionary of the dictionary storage unit 25.

Next, in step S36, the name registration unit 58 controls the echo back unit 56 to
Is output to the output control unit 57 (permission of execution of the process of step S15 in FIG. 8 is permitted). As a result, the echo back sound of the registered name is transmitted to the speaker 1 via the output control unit 57.
8 and output.

The echo back unit 56 determines in step S29
Since the execution of the processes in steps S11 to S14 of FIG. 8 is permitted, when a voice (name) is input in step S31, the memory 45 of the echo back unit 56 stores the echo back voice. Data is stored.

On the other hand, if the name registration unit 58 is notified in step S55 of FIG. 12 that the name is not appropriate, it determines in step S34 that it cannot be registered as a name, and proceeds to step S37, where the action determination mechanism unit The control unit 52 controls the posture transition mechanism unit 53 to output the action command information of the action indicating that “the spoken name was not registered”. Thereby, the posture transition mechanism unit 53 changes the posture of the robot to each posture in the behavior indicating that “the spoken name was not registered” based on the behavior command information from the behavior determination mechanism unit 52. And generates the posture transition information for transmission to the control mechanism unit 54. Control mechanism 54
Is based on the posture transition information from the posture transition mechanism 53.
Generates a control signal for driving the actuator 3AA ₁ to 5A ₁ and 5A _2, which, actuator 3
AA ₁ to be sent to 5A ₁ and 5A _2. This allows
Actuators 3AA ₁ to 5A ₁ and 5A ₂ is driven in accordance with the control signals, the robot take action indicating that "name that is spoken is not registered".

In step S27, when it is determined that 10 seconds have elapsed, that is, when the head has not been touched within 10 seconds after the action indicating "I want to touch my head" is performed, the flow proceeds to step S38. The name registration unit 58 controls the voice recognition unit 50A of the sensor input processing unit 50 to restart the voice recognition stopped in step S23. In addition,
At this time, the name registration unit 58 controls the echo back unit 56 to permit the execution of the echo back process (FIG. 8).

In step S32, when it is determined that 10 seconds have elapsed, that is, when no voice has been input within 10 seconds after the action indicating "I want you to give a name" has been performed, or in step S38. When the voice recognition is restarted, the process proceeds to step S39, where the name registration unit 58
Controls the action determination mechanism 52 to cause the attitude transition mechanism 53 to output action command information of an action indicating that “name has not been input”. Thereby, the posture transition mechanism unit 53 sets the posture of the robot to “the name has not been input” based on the action command information from the action determination mechanism unit 52.
Posture transition information for transitioning to each posture in the action indicating the fact is generated and transmitted to the control mechanism unit 54. Control mechanism unit 54 in accordance with the posture transition information from the attitude transition mechanism part 53, the actuator 3AA ₁ to 5A ₁ and 5A ₂
It generates a control signal for driving the, which is sent to the actuator 3AA ₁ to 5A ₁ and 5A _2. Thus, the actuator 3AA ₁ to 5A ₁ and 5A ₂
Is driven according to the control signal, and the robot takes an action indicating that “name has not been input”.

In step S36, when an echo back sound is output, in step S37, when an action indicating that "the spoken name has not been registered" is performed, or in step S39, "name is input." When the action indicating “not performed” has been performed, the process ends.

As described above, the voice input after the robot has performed an action indicating "I want you to give a name" is recognized as a name. Can be properly recognized as a name.

In the above description, a case has been described as an example where the robot starts name registration processing when the robot has grown to a state where a name is required. However, as shown in the flowchart of FIG. The voice recognition unit 50A of the sensor input processing unit 50 recognizes the instruction of “name registration”, and in step S62, notifies the name registration unit 58.
, The name registration process may be started. That is, in the case of this example, the user can give a name to the robot by speaking “name registration”.

In steps S63 to S79, the same processes as those in steps S23 to S39 in FIG. 9 are performed, and thus description thereof will be omitted.

Next, the procedure of another name registration process will be described with reference to the flowchart of FIG. In this case, as in the case of the example of FIG.
In 1, it is determined that the robot has grown to a state that requires a name, and is started when the fact is notified to the name registration unit 58 in step S92.

When the name registration unit 58 is notified that the robot has grown to a state requiring a name, the name registration unit 58 controls the voice recognition unit 50A of the sensor input processing unit 50 in step S93. Then, the voice recognition of the voice signal given from the microphone 15 is stopped. As a result, voice recognition is not performed until voice recognition is restarted in steps S98 and S109 described below. At this time, the name registration unit 58 controls the echo back unit 56 to prohibit the execution of the echo back process (FIG. 8).

In step S94, name registration unit 58
Controls the action determination mechanism 52 to send the action command information indicating “I want you to give a name” to the posture transition mechanism 53
Output. Thereby, based on the action command information from the action determining mechanism section 52, the attitude transition mechanism section 53 changes the attitude of the robot to each of the attitudes in the action indicating “I want you to give a name”. The information is generated and transmitted to the control mechanism 54. Control mechanism unit 54 in accordance with the posture transition information from the attitude transition mechanism part 53 generates control signals for driving the actuators 3AA ₁ to 5A ₁ and 5A _2, which, actuators 3AA ₁
To be sent to 5A ₁ and 5A _2.

Actuators 3AA _{1 to} 5A ₁ and 5
A ₂ is driven in accordance with the control signals, the robot, for example, as shown in FIG. 11, take action indicating that "I want to name" (robot, their ears, the above-friendly, it Shake left and right several times).

Next, in step S95, the name registration unit 58 resets and starts a timer T built in the controller 10.

In step S96, name registration unit 58
Determines whether or not the user has touched the head of the robot by communicating with the pressure processing unit 50C of the sensor input processing unit 50.

If it is determined in step S96 that the user has not touched the head of the robot, the flow advances to step S97, and the name registration unit 58 determines whether the value of the timer T started in step S95 is 10 or more. (If 10 seconds have elapsed), and if it is determined that 10 seconds have not elapsed, the process returns to step S96, and the subsequent processing is executed.

On the other hand, if it is determined in step S96 that the head has been touched, the flow advances to step S98, where the name registration unit 58 controls the voice recognition unit 50A of the sensor input processing unit 50, and stops in step S93. Then, the voice recognition is restarted. At this time, the name registration unit 58 controls the echo back unit 56 to permit execution of the processing from step S11 to step S14 in the processing shown in FIG. As a result, the data of the echo-back sound generated from the input sound is stored in the memory 45.

Next, in step S99, the name registration section 58 turns on the LED 19.

At step S100, the name registration unit 5
Step 8 resets and starts the timer T.

Next, in step S 101, the name registration unit 58 makes the voice recognition unit 50 A of the sensor input processing unit 50
By communicating with, it is determined whether or not a voice signal has been input to the voice recognition unit 50A.

If it is determined in step S101 that a voice signal is not input, the process proceeds to step S102, and the name registration unit 58 determines whether the value of the timer T started in step S100 is 10 (whether 10 seconds have elapsed). If it is determined that 10 seconds have not elapsed, the process returns to step S101, and the subsequent processing is executed.

If it is determined in step S101 that an audio signal has been input, the process proceeds to step S103, where the name registration unit 58 turns off the LED 19 turned on in step S99.

Steps S104 to S109,
And in step S111, step S111 in FIG.
Since the same processing as in the steps 33 to S39 is performed, the description thereof is omitted.

If it is determined in step S102 that 10 seconds have elapsed, the process proceeds to step S110, where the name registration unit 5
8 turns off the LED 19 turned on in step S99.

Also in the case of this example, as shown in the flowchart of FIG. 15, the voice recognition unit 50A of the sensor input processing unit 50 recognizes the command of “name registration” in step S121, and in step S122, When the notification to the effect is sent to the name registration unit 58, the name registration process may be started.

In steps S123 to S141, steps S93 to S11 in FIG.
Since the same processing as in the case of No. 1 is performed, the description thereof is omitted.

In the above description, a case has been described in which, after a name is registered, an echo-back sound generated from a sound representing the name is output, but a predetermined sound stored in the memory 45 in advance is output. It can also be done.

In addition, the memory 45 holds the data of the echo-back sound generated from the registered sound (name), and the user utters “What is the name?” And the robot recognizes it. At this time, the echo back sound having the name stored in the memory 45 may be output.

The case where the present invention is applied to an entertainment robot (robot as a pseudo pet) has been described above. However, the present invention is not limited to this, and may be applied to various robots such as industrial robots. It can be widely applied. In addition, the present invention is applicable not only to a robot in the real world but also to a virtual robot displayed on a display device such as a liquid crystal display.

Furthermore, in the present embodiment, the above-described series of processing is performed by causing the CPU 10A to execute a program, but the series of processing may be performed by dedicated hardware. .

The program is stored in the memory 1 in advance.
0B (FIG. 2), a floppy (registered trademark) disk, a CD-ROM (Compact Disc Read Only Memor
y), MO (Magneto optical) disc, DVD (Digital Versa)
It can be temporarily or permanently stored (recorded) in a removable recording medium such as a tile disc), a magnetic disk, or a semiconductor memory. Then, such a removable recording medium can be provided as so-called package software, and can be installed in the robot (memory 10B).

The program is transferred from a download site wirelessly via an artificial satellite for digital satellite broadcasting, or by wire via a network such as a LAN (Local Area Network) or the Internet. Can be installed.

In this case, when the program is upgraded, the upgraded program can be easily installed in the memory 10B.

Here, in this specification, the CPU 10A
The processing steps for writing a program for causing the CPU to perform various types of processing do not necessarily need to be processed in chronological order in the order described in the flowchart, and may be performed in parallel or individually (for example, parallel processing). Or processing by an object).

Further, the program may be processed by one CPU or may be processed by a plurality of CPUs in a distributed manner.

[0151]

The robot control apparatus and method of the present invention,
In addition, according to the program on the recording medium, the robot controls the behavior of the robot so as to cause the robot to perform an action indicating that a name is to be given. Since the detection and the registration of the optimal phoneme string are performed as names, the names can be registered appropriately.

[Brief description of the drawings]

FIG. 1 is a perspective view illustrating an external configuration example of a robot according to an embodiment of the present invention.

FIG. 2 is a block diagram illustrating an example of an internal configuration of a robot.

FIG. 3 is a block diagram illustrating a functional configuration example of a controller 10;

FIG. 4 is a diagram showing a growth model.

FIG. 5 is a block diagram illustrating a configuration example of a voice recognition unit 50A.

FIG. 6 is a diagram showing a word dictionary.

FIG. 7 is a block diagram illustrating a configuration example of an echo back unit 56.

FIG. 8 is a flowchart illustrating an echo back process performed by an echo back unit.

FIG. 9 is a flowchart illustrating a name registration process.

FIG. 10 is a diagram illustrating the behavior of the robot.

FIG. 11 is a diagram illustrating another action of the robot.

FIG. 12 is a flowchart illustrating details of a process in step S33 of FIG. 9;

FIG. 13 is a flowchart illustrating another name registration process.

FIG. 14 is a flowchart illustrating another name registration process.

FIG. 15 is a flowchart illustrating another name registration process.

[Explanation of symbols]

1 head unit, 4A lower jaw, 10 controller, 10A CPU, 10B memory, 15
Microphone, 16 CCD camera, 17 touch sensor, 18 speaker, 19 LED, 21 AD
Conversion unit, 22 feature extraction unit, 23 matching unit,
24 acoustic model storage unit, 25 dictionary storage unit, 26
Grammar storage unit, 27 voice section detection unit, 31 text generation unit, 32 rule synthesis unit, 34 dictionary storage unit, 35 grammar storage unit for generation, 36 phoneme segment storage unit,
41 AD conversion unit, 42 prosody analysis unit, 43 sound generation unit, 44 output unit, 45 memory, 46 voice section detection unit, 50 sensor input processing unit, 50A
Voice recognition section, 50B image recognition section, 50C pressure processing section, 51 model storage section, 52 action determination mechanism section, 53 attitude transition mechanism section, 54 control mechanism section,
55 voice synthesis unit, 56 echo back unit, 57 output control unit, 58 name registration unit

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G10L 3/00 551H (72) Inventor Jun Hiroi 6-7-35 Kita Shinagawa, Shinagawa-ku, Tokyo Sony Stock In-house (72) Inventor Wataru Onoki 6-7-35 Kita-Shinagawa, Shinagawa-ku, Tokyo F-term in Sony Corporation (reference) 2C150 CA02 DA05 DA24 DA26 DA27 DA28 DF03 DF04 DF33 ED42 ED52 EF07 EF16 EF23 EF29 EF33 EF36 3F059 AA00 BA00 BB06 DA05 DC00 FC00 3F060 AA00 BA10 CA14 5D015 GG03 KK02 LL10 LL11

Claims (6)

[Claims] 1. A robot control device for controlling a robot that behaves based on at least a speech recognition result, wherein the behavior control means controls the behavior of the robot so that the robot performs a behavior indicating that a name is desired. Detecting means for detecting an optimum phoneme string from a voice input after the action of the robot is controlled by the action control means; and registration means for registering the optimum phoneme string as the name. A robot control device characterized by the following. 2. The robot according to claim 1, further comprising a transition unit configured to sequentially transition a growth state of the robot to a predetermined state, wherein the behavior control unit transitions the growth state of the robot to a predetermined state by the transition unit. The robot control device according to claim 1, wherein the robot controller controls the behavior of the robot so that the robot performs an action indicating that the user wants to give the name. 3. An acquisition unit for acquiring a command input by a user for instructing start of a name registration process, wherein the action control unit is configured to, when the command is acquired by the acquisition unit, enter the name. The robot control device according to claim 1, wherein the control unit controls the behavior of the robot so that the robot performs an action indicating that the user wants to attach a character. 4. The method according to claim 1, further comprising: first storage means for storing an acoustic model representing an acoustic feature; and second storage means for storing a word dictionary in which words for voice recognition are registered. Means is a word model consisting of the acoustic model connected corresponding to the feature parameter of the voice,
Detecting the word model having the highest score in which the feature parameter is observed as the optimal phoneme string, wherein the registration unit is connected in correspondence with the phoneme of the word already registered in the word dictionary. Among the word models composed of acoustic models, the word model having the highest score at which the feature parameter is observed is detected, and the difference between the detected word model score and the score of the optimal phoneme sequence is determined by a predetermined value. The robot control device according to claim 1, wherein when the value is larger than a threshold value, the optimal phoneme sequence is registered as the name. 5. A robot control method for a robot control device for controlling a robot acting based on at least a speech recognition result, wherein the robot controls an action of the robot so that the robot performs an action indicating that a name is desired. A control step of detecting an optimal phoneme sequence from a voice input after the behavior of the robot is controlled by the behavior control step; and a registration step of registering the optimal phoneme sequence as the name. A robot control method comprising: 6. A program for a robot control device for controlling a robot acting at least based on a speech recognition result, wherein the program controls the behavior of the robot so that the robot performs an action indicating that a name is desired. Behavior control means, detection means for detecting an optimal phoneme sequence from a voice input after the behavior control means controls the behavior of the robot, and registration means for registering the optimal phoneme sequence as the name. A recording medium on which a computer-readable program is recorded.