JP4284989B2 - Speech analyzer - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
  The present invention analyzes animal speechSpeech analyzerAbout.
[0002]
[Prior art]
For those who keep animals such as dogs and cats as pets, pets are members of the family, and there is a desire to communicate and communicate with each other in the same way as humans.
[0003]
In recent years, with advances in speech analysis technology, particularly voiceprint analysis, it has become possible to discriminate emotions and intentions included in animal calls (hereinafter simply referred to as “emotions”). For example, a voice (such as a sonograph) obtained by voice analysis of voices generated by animals such as pets and domestic animals is obtained. Then, the emotion of the animal is discriminated by comparison with a reference voice pattern prepared in advance and analyzed in terms of animal behavior.
[0004]
Based on such voice analysis technology, for example, input the voice of the animal's cry and the video of the animal's movement (gesture) and compare it with the voice and movement data analyzed in advance in animal behavior It has also been proposed to discriminate animal emotions and display them as characters or images that can be understood by humans (see, for example, Patent Document 1; corresponding to all claims).
[0005]
[Patent Document 1]
Japanese Patent Laid-Open No. 10-3479
[0006]
[Problems to be solved by the invention]
According to such a technique, the owner can know the emotion of the animal to some extent, and when there is a request from the animal, it can understand and respond to it. However, the transmission of emotions realized by the prior art is only one way from animals to humans, and the transmission of emotions from humans to animals is not supported. Therefore, it was difficult to say that the owner and the pet were communicating.
[0007]
  The present invention has been made in view of the above-mentioned problems, and the object of the present invention is humans and animals.WithTo support two-way emotional communication.
[0008]
[Means for Solving the Problems]
  In order to solve the above-mentioned problem, the speech analysis apparatus according to claim 1Has the following configuration.
[0011]
As used herein, “animal voice” means crying. “Human language” means text or an image such as a human voice or a word that allows a human to understand its meaning. The term “animal language” means a sound pattern that enables communication within the same genus or group.
[0022]
  Claim1The speech analysis apparatus according to the invention described in
  In the voice analysis device provided with the voice analysis device body,
  First voice input means for inputting animal voice (for example, voice input unit 10 in FIG. 3, CPU 20, step S102 in FIG. 12);
  A first voice analysis means for analyzing the voice input by the first voice input means (for example, the voice analysis unit 12 of FIG. 3, the CPU 20, step S108 of FIG. 12, steps S202 to S216 of FIG. 13);
  First output means (for example, the sound output unit 50 of FIG. 3, the CPU 20, step S116 of FIG. 12, step S304 of FIG. 14), which displays and outputs characters or images corresponding to the analysis result by the first voice analysis means as human language. , S306)
  After the human language is displayed and output by the first output means, a second voice input means (for example, the voice input unit 10 in FIG. CPU 20, step S102 in FIG. 12,
  A second voice analysis means for analyzing the voice input by the second voice input means (for example, the voice analysis unit 12 of FIG. 3, the CPU 20, step S108 of FIG. 12, steps S202 to S216 of FIG. 13);
  The second output means (for example, the sound output unit 50 of FIG. 3, the CPU 20, step S112 of FIG. 12, FIG. 15) outputs the voice corresponding to the analysis result by the second voice analysis means as an animal language understandable by the animal. Steps S404 and S406),
  First registration means for registering a user's voiceprint (for example, ROM 40 in FIG. 3, voiceprint data 410 in FIG. 8);
  Second registration means (for example, ROM 40 in FIG. 3, human language / animal language conversion TBL 418 in FIG. 8, registered voice data 418f in FIG. 10) for registering speech in human language having a predetermined meaning, and the second voice input means Determining means (for example, the control unit 120 in FIG. 1, the CPU 20 in FIG. 3, and the steps S416 to S418 in FIG. 15) for determining whether or not the voice input by the first registration means coincides with the voiceprint registered by the first registration means. )When,
  Fourth output means (for example, the speaker unit 104 in FIG. 1, the CPU 20 in FIG. 3, the sound output) that outputs the voice registered by the second registration means in human language when it is determined by the determination means that they do not match. Unit 50, (steps S418 to S420, S424) of FIG.
  It is characterized by providing.
[0023]
  Claim1In the invention described inAccording to the userIs different from a pre-registered voiceprint, it is possible to output a pre-registered human voice in human language. Therefore, it is possible to obtain a training effect that allows the animal to hear and hear the human voice corresponding to the emotion of the user. In addition, when the voiceprint is the voiceprint of the person who is most familiar with the animal (for example, the owner) and the voice uttered by the person is registered by the second voice registration means, if the user is different from the person, By listening to the voice of the most familiar person, the effect of calming the animal can be obtained.
[0024]
  Claim2The invention described in claim1The voice analyzer describedIn the voice analyzer main bodyIt is characterized by comprising mounting means (for example, the wristband unit 110 in FIG. 1 and the clip unit 212 in FIG. 24) for mounting on the user's body.
[0025]
  Claim2According to the invention described in claim1While having the same effect as the described invention, the voice analyzer can be worn on the body and carried. Therefore, it is not necessary to take out from the back or the like at the time of using the speech analysis apparatus, and the usability is improved.
[0026]
  Claim3The invention described in claim1 or claim 2In the voice analysis device describedLeaveStorage means for storing the voice input by the first voice input means and the time when the voice is input by the first voice input means (for example, the CPU 20, RAM 30, FIG. 7 history data in FIG. 7) 322, step S117 of FIG. 12).
[0027]
  Claims4Invention described inIs, Claims3In the voice analyzer described inLeaveAnd fifth output means (for example, the monitor unit 106 in FIG. 1, the CPU 20 in FIG. 3, the display unit 52, and step S128 in FIG. 12) for outputting the voice and time stored in the storage unit in association with each other.It is characterized by.
[0028]
  Claim3According to the invention described in (1), it is possible to store a history of exchanges between animals and humans.
  Claim4Invention described inAccording toBy outputting a history of exchanges between animals and humans, the history can be analyzed and used.
[0029]
  Claim5The invention described in claim 14The voice analysis device according to any one ofInExcitation means (for example, the vibrator 112 in FIG. 1, the CPU 20 in FIG. 3, the excitation unit 54, and the vibration pattern TBL 420 in FIG. 11) that generates a predetermined vibration corresponding to the analysis result by the first voice analysis means is provided. It is characterized by providing.
[0030]
  Claim5According to the invention described in claim 1,4In addition to the same effects as the invention described in any one of the above, it is possible to generate a predetermined vibration corresponding to the analysis result by the first voice analysis means and notify the user of the analysis result by bodily sensation. Therefore, the user does not need to read the human language text or the like on the analysis result, so that the usability is improved and smoother communication can be achieved. Moreover, it becomes possible to know the analysis result even for a visually handicapped person or a hearing handicapped person.
[0031]
DETAILED DESCRIPTION OF THE INVENTION
[First Embodiment]
Next, a first embodiment of a speech analyzer to which the present invention is applied will be described with reference to FIGS. In this embodiment, the animal is described as a dog, but the present invention is not limited to this. For example, other animals such as cats, dolphins, and parrots may be used.
[0032]
[Description of configuration]
FIG. 1 is a diagram showing an example of the appearance of a wristwatch type speech analyzer to which the present invention is applied. As shown in FIG. 1 (a), the external appearance of the wristwatch type speech analysis apparatus 100 has the same form as a conventional wristwatch as a whole. Then, a microphone unit 102 for inputting animal and human voices, a speaker unit 104 for outputting sound, a monitor unit 106 for displaying and outputting text and images, a key operation unit 108 for inputting various operations, and the arm A wristband unit 110 for wearing on a human body or the like when carrying the hour meter voice analyzer 100, a vibrator 112, a data communication unit 114 for performing wireless communication with an external device, and a wristwatch type voice analyzer 100. A control unit 120 for integrated control and a power supply unit (not shown) are provided.
[0033]
The microphone unit 102 is a sound collection device and is realized by, for example, a microphone. In the figure, a single unit is provided, but there may be a plurality of units, or a configuration in which a cable that is detachably separated and connected to the main body can be attached by a clip or the like.
[0034]
The speaker unit 104 is a sound output device and is realized by, for example, a speaker. In this embodiment, since there is a case where high-frequency sound outside the audible range for humans is output, the speaker unit 104 is configured to output sound in the high-frequency range.
[0035]
The monitor unit 106 is a display output unit realized by a display element such as an LCD (Liquid Crystal Display) and an ELD (Electronic Luminescent Display), a backlight, a driving circuit, and the like. The monitor unit 106 can display characters (text), figures, images, and the like under the control of the control unit 120. In the figure, the monitor unit 106 is singular, but may be configured to include a plurality.
[0036]
The key operation unit 108 is input means realized by, for example, a button switch, a lever, a dial, or the like. In the present embodiment, as shown in FIG. 1B, an up direction key 108a, a down direction key 108c, a selection key 108b, and a cancel key 108d are provided. For example, a selection operation from a plurality of menus, a determination and cancellation operation, a calling operation of a predetermined function, and the like can be input depending on a combination of a key pressing time and a pressing order. The number of key operation units 108 is not limited to the above, and may be set as appropriate.
[0037]
The wristband unit 110 is a means for the user to attach and equip to the body and belongings when carrying it. For example, in addition to the wristband of a wristwatch, a clip, string, chain, magic tape ( Registered trademark), a magnet, or the like.
[0038]
Vibrator 112 is a small vibration device. In the present embodiment, the vibration is generated in a pattern corresponding to the emotion included in the voice of the dog 2 under the control of the control unit 120. By experiencing various vibration patterns, the user 4 can know the emotion and intention of the dog 2 without looking at the monitor unit 106, and can also be used by visually impaired persons and hearing impaired persons.
[0039]
The data communication unit 114 realizes data transmission / reception by wireless communication with an external device such as a personal computer. For example, a communication module corresponding to a standard such as Bluetooth (registered trademark) or IrDA, This is realized by a jack terminal or the like.
[0040]
The control unit 120 includes a CPU (Central Processing Unit), various IC memories, a crystal oscillator, and the like, and the wristwatch type speech analysis apparatus 100 is integrated by the CPU reading and processing programs stored in the IC memory. Can be controlled. Further, for example, the wristwatch type voice analysis apparatus 100 can be made to function as a wristwatch using a crystal oscillator or the like.
[0041]
FIG. 2 is a conceptual diagram showing an example of a usage method in the present embodiment. As shown in the figure, the user 4 uses the wristwatch type voice analysis device 100 by wearing it on the wristband unit 110, for example, on the wrist of the user 4 and carrying it. Carrying the wristwatch-type speech analysis apparatus 100 like a wristwatch can eliminate the inconvenience of separately carrying the apparatus and the inconvenience of taking out the apparatus from the bag at the time of use.
[0042]
The wristwatch type voice analysis apparatus 100 captures (detects) the voice of the user 4 and the pet dog 2 on the user 4 side, and supports bidirectional communication between the two. That is, when the voice of the dog 2 is captured by the microphone unit 102, the voice of the dog 2 is analyzed to discriminate the emotion of the dog 2, and the monitor unit 106 generates a text or a figure (human language) that can be understood by the user 4. indicate. On the other hand, when the voice of the user 4 is captured, voice analysis is performed to determine the emotion of the user 4 and the speaker unit 104 outputs a sound (animal language) that can be understood by the dog 2.
[0043]
[Description of functional block]
FIG. 3 is a functional block diagram illustrating an example of a functional configuration according to the present embodiment.
As shown in the figure, the wristwatch type voice analysis apparatus 100 includes a voice input unit 10, a voice analysis unit 12, a key input unit 14, a voice analysis ROM (Read Only Memory) 16, a CPU 20, and a RAM ( Random Access Memory) 30, ROM 40, sound output unit 50, display unit 52, vibration unit 54, communication unit 60, and system bus 90.
[0044]
The voice input unit 10 inputs the voice of the dog 2 or the user 4 and outputs a voice signal to the voice analysis unit 12. In FIG. 1, the microphone unit 102 corresponds to this.
[0045]
The voice analysis unit 12 performs voice analysis on the voice signal input from the voice input unit 10. More specifically, for example, a process for removing a noise component included in an audio signal, a process for converting an audio signal into A / D conversion into audio data of a predetermined format, a patterning process for extracting features of audio data, Comparison processing with a reference voice pattern registered in advance is executed. These processes can be realized by an arithmetic processing integrated circuit such as an A / D converter, a filter circuit, or a DSP (Digital Signal Processor), for example. A configuration in which some or all of the functions are realized by reading a program or data stored in the voice analysis ROM 16 and performing arithmetic processing (implemented by software) may be employed. The voice analysis unit 12 is mounted on the control unit 120 in FIG.
[0046]
The voice analysis ROM 16 stores programs and data used for various processes of the voice analysis unit 12 and is referred to by the voice analysis unit 12. In FIG. 1, the voice analysis ROM 16 is mounted on the control unit 120.
[0047]
FIG. 4 is a diagram showing an example of the contents stored in the voice analysis ROM 16 in the present embodiment. As shown in the figure, for example, a voice analysis program 162 that is a program for realizing various processes of the voice analysis unit 12 by arithmetic processing, and data serving as a reference for comparison with the voice input from the voice input unit 10. A certain animal reference voice pattern 164 and a human reference voice pattern 166 are stored.
[0048]
FIG. 5 is a diagram showing an example of the data configuration of the animal reference voice pattern 164 in the present embodiment. As shown in the figure, the animal reference voice pattern 164 is prepared in advance for each type of animal (animal attribute code), and includes an animal attribute code 164a indicating the animal species to be applied, and information for classifying animal emotions. A certain emotion identification code 164b and a reference voice pattern 164c of a voice (scream) corresponding to an animal word for transmitting this emotion are stored in association with each other. The reference voice pattern 164c is, for example, sonograph data. Animal language means a sound pattern that enables communication within the same genus and group.
[0049]
The animal reference voice pattern 164 is information obtained by a statistical method and analyzed in terms of animal behavior. Based on the animal attribute code 164a, the animal reference voice pattern 164 matching the target animal is searched, the voice data of the voice input from the voice input unit 10 is patterned, and matching with the reference voice pattern 164c is determined. Thus, it is possible to discriminate the emotions of animals included in this voice.
[0050]
The human reference voice pattern 166 is information serving as a reference for discriminating emotions included in the voice of the user 4, and is prepared in advance according to the applied human attribute. The human attribute mentioned here is a classification using, for example, language classification, gender, age, and the like as parameters.
[0051]
For example, as shown in FIG. 6, the human reference voice pattern 166 includes a human attribute code 166a indicating a human attribute to be applied, an emotion identification code 166b for classifying a human emotion, and a human voice reference corresponding thereto. Audio pattern 166c.
[0052]
The reference voice pattern 166c is a voice pattern when a characteristic voice pattern statistically obtained and analyzed or a word expressing emotion is pronounced, and is, for example, sonograph data. Therefore, by searching for the human reference voice pattern 166 of the human attribute code 166a that matches the user 4, patterning the voice data of the voice input from the voice input unit 10, and determining the matching with the reference voice pattern 166c, The emotion of the user 4 included in this voice can be determined. Note that the data included in the human reference voice pattern 166 is not limited to the above, and includes data necessary for discrimination, such as judgment values such as the speed of pronunciation of words and strength of speech, as appropriate, and may be used for discrimination of matching. I do not care.
[0053]
The key input unit 14 is realized by, for example, a button switch, a lever, a dial, a touch panel, a track pad, and the like, and inputs an operation and outputs an operation signal to the CPU 20. In FIG. 1, the key operation unit 108 corresponds to this.
[0054]
The CPU 20 is mounted on the control unit 120 in FIG. 1 and executes various processes by controlling each block in an integrated manner through arithmetic processing.
[0055]
The RAM 30 is an IC memory in which the CPU 20 and the voice analysis unit 12 temporarily store programs and data, and is mounted on the control unit 120 in FIG.
[0056]
FIG. 7 is a diagram showing an example of contents stored in the RAM 30 in the present embodiment. As shown in the figure, for example, an animal name 302 that stores name information of the dog 2, an animal attribute code 304, a human attribute code 306, time data 308, audio data 310, and audio input time data 312 The voice identification flag 314, the emotion identification code 316, the high frequency mode flag 318, the sensation mode flag 320, and the history data 322 are stored.
[0057]
The animal name 302 is information indicating the name of the dog 2, and the animal attribute code 304 is information indicating the type of the dog 2. Both are registered by the user 4 before use. The animal name 302 is displayed on the monitor unit 106 in a human language output process or the like, which will be described later, and functions to increase the intimacy between the dog 2 and the user 4.
[0058]
The human attribute code 306 is information indicating an attribute (for example, language type, sex, age, etc.) of the user 4 and is registered by the user 4 before use.
[0059]
The time measurement data 308 is information indicating date and time information. By referring to the time measurement data 308, the wristwatch type speech analysis apparatus 100 can also function as a clock or a timer.
[0060]
The voice data 310 is digital data obtained by converting the voice input from the voice input unit 10 by the voice analysis unit 12. In the present embodiment, the waveform data is stored, but other data formats such as a sonograph may be used. The time when the voice that is the source of the voice data 310 is input is stored in the voice input time data 312.
[0061]
The voice identification flag 314 and the emotion identification code 316 store the result of voice analysis of the voice data 310 by the voice analysis unit 12. The voice identification flag 314 is information indicating whether the voice data is animal voice or human voice. The emotion identification code 316 stores the emotion identification code 164b or 166b determined by matching with the reference voice pattern 164c or 166c.
[0062]
The high-frequency mode flag 318 outputs the high-frequency sound that can be heard by the dog 2 but is not heard by humans when the emotion of the user 4 is determined and the sound is output from the speaker unit 104 using the animal language in the animal language output process described later. This is information for setting whether or not to perform. For example, when the animal is a dog, the high-frequency sound corresponds to a sound in a range produced by a so-called “dog flute”.
[0063]
The bodily sensation mode flag 320 discriminates emotions included in the voice of the dog 2 in the human language output process described later, and causes the vibrator 112 to vibrate when displaying text or a figure that can be understood by the user 4 on the monitor unit 106. This is information for setting whether or not to generate.
[0064]
The history data 322 is a history relating to voice input and output, and stores a voice input time 322a, a voice identification flag 322b, and an emotion identification code 322c in association with each other. Accordingly, by referring to the history data 322, it is possible to know when and who (the dog 2 or the user 4) showed what emotions were exchanged.
[0065]
The ROM 40 stores programs and data for causing the CPU 20 to realize various functions through arithmetic processing.
[0066]
FIG. 8 is a diagram showing an example of contents stored in the ROM 40 in the present embodiment. As shown in the figure, as a program, a system program 400 and a text or figure (human language) that can be understood by the user 4 (human) based on the voice analysis result of the voice of the dog 2 (animal) are output. A human language output program 402 for executing a human language output process and an animal language output process for outputting a sound that can be understood by the dog 2 based on a voice analysis result of the voice of the user 4 and the like. A program 404, a mode switching program 406 for executing processing relating to various mode switching, and a history output program 408 for executing history display processing based on the history data 322 are included.
[0067]
As data, voice print data 410 used for user 4 identification, clock display data 412 for displaying a clock on the monitor 106, screen frame data 414 for storing information necessary for various screen displays, animal data A language / language conversion TBL (table) 416, a language / animal language conversion TBL (table) 418, and a vibration pattern TBL (table) 420 are stored.
[0068]
The voiceprint data 410 is a voiceprint of a person familiar to the dog 2 on a daily basis, for example, a voiceprint of an owner, and is collected and stored in advance, for example, by a manufacturer of a wristwatch type voice analyzer. Of course, the voiceprint data 410 is not limited to be stored in the ROM 40, but may be configured to be registered in the RAM 30 by the user 4.
[0069]
The animal language human language conversion TBL 416 stores the emotion of the dog 2 in association with the human language, and the human language animal language conversion TBL 418 corresponds to dictionary data that stores the emotion of the user 4 in association with the animal language. Information.
[0070]
FIG. 9 is a diagram illustrating an example of a data configuration of the animal language human language conversion TBL 416 according to the present embodiment. As shown in the figure, the animal language human language conversion TBL 416 includes an emotion identification code 416a that the voice analysis unit 12 analyzed and discriminated from the voice of the dog 2, text data 416b that can be understood by humans, and an animal The image data 416c for displaying the image is stored in association with each other. The image data 416c may be still image information or moving image information for displaying an animation or the like.
[0071]
FIG. 10 is a diagram illustrating an example of a data configuration of the human-animal language conversion TBL 418 according to the present embodiment. As shown in the figure, the human-animal language conversion TBL 418 includes an emotion identification code 418a determined by the voice analysis unit 12 analyzing and analyzing the voice of the user 4, and corresponding human-readable text data 418b. Image data 418c for displaying a human image, synthesized sound data 418d obtained by artificially synthesizing a voice of an animal (in this case, a dog), high-frequency sound data 418e outside the human audible range, and pre-registered The registered voice data 418f, which is the voice of the user 4, is stored in association with it. The image data 418c may be still image information or moving image information for displaying an animation or the like.
[0072]
For example, as shown in FIG. 11, the vibration pattern TBL 420 stores an emotion identification code 420 a and a vibration pattern 420 b in association with each other. By referring to the vibration pattern TBL420, the vibrator 112 can be vibrated with the vibration pattern 420b corresponding to the emotion identification code 420a.
[0073]
The sound output unit 50 is realized by, for example, a speaker and outputs sound. The speaker unit 104 in FIG. 1 corresponds to this.
[0074]
The display unit 52 is realized by, for example, a display element such as an LCD, ELD, or PDP, and displays and outputs an image. The monitor unit 106 in FIG. 1 corresponds to this.
[0075]
The vibration unit 54 is realized by a vibrator such as a vibrator, for example, and generates vibration. The vibrator 112 in FIG. 1 corresponds to this.
[0076]
The communication unit 60 is transmission / reception means for performing wireless communication with an external device. For example, it is realized by a module such as Bluetooth (registered trademark) or IrDA, a jack of a wired communication cable, a control circuit, or the like. The data communication unit 114 in FIG. 1 corresponds to this. Note that information such as a protocol stack provided by the communication unit 60 during communication is stored in the ROM 40 (not shown), and is appropriately read and used.
[0077]
[Description of processing]
Next, the flow of processing in the present embodiment will be described with reference to FIGS.
[0078]
FIG. 12 is a flowchart for explaining the main processing flow in the present embodiment. As shown in the figure, if the voice input unit 10 detects voice input (step S102), the voice analysis unit 12 performs A / D conversion and filter processing on the voice signal input from the voice input unit 10. Then, the voice data 310 is converted into an appropriate format suitable for voice analysis (step S104).
Next, the timing data 308 at that time is stored as voice input time data 312 in association with the voice data 310 (step S106), and voice analysis processing of the voice data 310 is executed (step S108).
[0079]
FIG. 13 is a flowchart for explaining the flow of the voice analysis processing in the present embodiment. As shown in the figure, the voice analysis unit 12 first reads the stored voice data 310 (step S202) and executes matching with the animal reference voice pattern 164 (step S204). That is, the sonographic data is obtained by patterning the audio data 310, and compared with the pattern of the reference audio pattern 164c. If there is an approximation of the pattern feature, it is determined that there is a match.
[0080]
If there is a match in the animal reference voice pattern 164 (step S206; YES), the voice analysis unit 12 stores “1” indicating the voice of the animal in the voice identification flag 314 and matches the matched reference voice pattern. The emotion identification code 164b corresponding to 164c is stored in the emotion identification code 316 of the RAM 30 (step S208), the voice analysis process is terminated, and the flow returns to the flow of FIG.
[0081]
If there is no matching in the animal reference voice pattern 164 (step S206; NO), matching with the human reference voice pattern 166 is executed (step S210).
[0082]
If there is a matching in the human reference voice pattern 166 (step S212; YES), the voice analysis unit 12 stores “0” indicating human voice in the voice identification flag 314, and the matched reference voice pattern. The emotion identification code 166b corresponding to 166c is stored in the emotion identification code 316 of the RAM 30 (step S214), the voice analysis process is terminated, and the flow returns to the flow of FIG.
[0083]
If there is no matching in the human reference voice pattern 166 (step S212; NO), the voice analysis unit 12 stores “0” in the voice identification flag 314 and “0” in the emotion identification code 316 of the RAM 30. Store (step S216), end the voice analysis process, and return to the flow of FIG.
[0084]
When the voice analysis process ends and the flow returns to the flow of FIG. 12, the CPU 20 refers to the voice identification flag 314 and the emotion identification code 316.
When the voice identification flag 314 is “1”, that is, when the voice of the dog 2 which is an animal is input (step S110; YES), the human language output process is executed (step S112). When the voice identification flag 314 is “0”, that is, when the voice of the user 4 who is a human is input (step S114; YES), the animal language output process is executed (step S116). When the voice identification flag 314 is “0” and the emotion identification code 316 is “0”, that is, when neither the animal voice nor the human voice can be discriminated (step S114; NO), the human language output process Does not move to animal language output processing.
[0085]
FIG. 14 is a flowchart for explaining the flow of the human language output process in the present embodiment. As shown in the figure, the CPU 20 first refers to the screen frame data 414 and causes the display unit 52 to display a frame for human language output (step S302).
[0086]
Next, the emotion identification code 316 in the RAM 30 is referred to (step S304), the text data 416b and the image data 416c corresponding to the emotion identification code 316 are read from the animal language human language conversion TBL 416, and the human language output screen is displayed. It is displayed at a predetermined position (step S306).
[0087]
Next, the voice data 310 is read out, the voice data is displayed at a predetermined position in the human language output screen (step S308), the voice input time data 312 is read out, and the date and time when the voice is input is displayed (step S308). S310).
[0088]
Next, the CPU 20 refers to the sensation mode flag 320, and when the sensation mode flag is “1”, that is, when the sensation mode is set to “ON” (step S312; YES), the vibration pattern TBL420 starts. The vibration pattern 420b corresponding to the emotion identification code 316 read out is read out. And the vibration part 54 is controlled according to the read vibration pattern 420b, a vibration is generated (step S314), a human language output process is complete | finished, and it returns to the flow of FIG. Then, when returning to the flow of FIG. 12, the CPU 20 updates the history data 322 (step S117).
[0089]
FIG. 18 is a diagram illustrating an example of a screen in the human language output process according to the present embodiment. In the human language output screen 5, the title display 5a indicates a message from the dog 2 to the user 4. At this time, for example, the user 4 is given a sense of intimacy by including the animal name 302 (the name of the pet) in the display like “a message from (Taro) has arrived”.
[0090]
Text data 416b and image data 416c corresponding to the emotion identification code 316 read from the animal language human language conversion TBL 416 are displayed on the text display unit 5b and the image display unit 5c, respectively. For example, the text display unit 5b is more preferably displayed in a balloon from the image display unit 5c.
[0091]
The audio data 310 is displayed in a graph on the audio data display unit 5d. It may be displayed as waveform data, or may be displayed in other formats such as a sonograph. By displaying the voice data 310 here, the user 4 develops the sense of reading the characteristics (graph shape, etc.) of the display, and eventually the graph of the voice data 310 without reading the text on the text display unit 5b. You can understand the feelings and intentions of Dog 2 just by looking at the display. The shape of this graph includes finer emotions and intentions. When the user 4 has a sense of reading the characteristics of the graph, the user can understand the dog 2 more finely than classification by the emotion identification code. become able to.
The date and time display unit 5e displays the time when the voice is input, for example, at the bottom of the screen.
[0092]
FIG. 15 is a flowchart for explaining the flow of the animal language output process according to the present embodiment. As shown in the figure, the CPU 20 first refers to the screen frame data 414 and displays a frame for animal language output on the display unit 52 (step S402).
[0093]
Next, the emotion identification code 316 in the RAM 30 is referred to (step S404), and the text data 418b and the image data 418c corresponding to the emotion identification code 316 are read from the human-animal language conversion TBL 418, and the animal language output screen is displayed. It is displayed at a predetermined position (step S406).
[0094]
Next, the voice data 310 is read, the voice data is displayed in a graph at a predetermined position in the animal language output screen (step S408), and the voice input time data 312 is read to display the date and time when the voice was input ( Step S410).
[0095]
Next, the CPU 20 refers to the high frequency mode flag 318, and when the high frequency mode flag is “1”, that is, when the high frequency mode is set to “ON” (step S412; YES), the human-animal language conversion TBL418. Thus, the high frequency sound data 418e corresponding to the previously referred emotion identification code 316 is read out and output from the sound output unit 50 (step S414).
[0096]
Next, the voice data 310 is checked against the voiceprint data 410 (step S416), and it is determined whether or not they match (step S418).
[0097]
When it is determined that the voice data 310 matches the voiceprint data 410 and is the voice of the person (step S418; YES), the synthesized sound data 418d corresponding to the emotion identification code 316 is read from the human-animal language conversion TBL418 (step S418). S422), output from the sound output unit 50 (step S424).
[0098]
When it is determined that the voice data 310 does not match the voiceprint data 410 (step S418; NO), the registered voice data 418f corresponding to the emotion identification code 316 is read from the human-animal language conversion TBL418 (step S420), and the sound output unit 50 (step S424). By outputting the registered voice data 418f, when the person who is not the owner is the user 4, by listening to the voice of a person familiar to him / her, the dog 2 is relieved of tension and alertness. Even if you are not familiar with, make communication more smoothly.
[0099]
If the synthesized sound data 418d or the registered voice data 418f is output from the sound output unit 50, the animal language output process is terminated and the flow returns to the flow of FIG. Then, when returning to the flow of FIG. 12, the CPU 20 updates the history data 322 (step S117).
[0100]
FIG. 19 is a diagram illustrating an example of a screen in the animal language output process according to the present embodiment. On the animal language output screen 6, the title display 6a indicates a message to the dog 2. At this time, for example, the user 4 is given a sense of intimacy by including the animal name 302 (the name of the pet) in the display like “I will input a message to (Taro)”.
[0101]
Text data 418b and image data 418c corresponding to the emotion identification code 316 read out from the human-animal language conversion TBL 418 are displayed on the text display unit 6b and the image display unit 6c, respectively. It is more preferable that the text display unit 6b is displayed in, for example, a balloon from the image display unit 6c as shown in FIG.
[0102]
The voice data 310 is displayed as a graph on the voice data display unit 6d, and the time when the voice is input is displayed on the date / time display unit 6e at the bottom of the screen.
[0103]
In the flow of FIG. 12, for example, when the up key 108a or the down key 108c is pressed for a relatively long time (step S118; YES), the CPU 20 executes a key input process (step S120).
[0104]
FIG. 16 is a flowchart for explaining the flow of key input processing in the present embodiment. As shown in the figure, the CPU 20 first refers to the screen frame data 414 to display a key input frame on the display unit 52 (step S502). On the key input screen, for example, referring to the human-animal language conversion TBL 418, the contents of the text data 418b are displayed in a selectable button (step S116).
[0105]
The user 4 selects a button having a desired content with the up direction key 108a or the down direction key 108c, and presses the selection key 108b to select and determine (step S504).
[0106]
If selection decision is input, CPU20 will select the emotion identification code 418a corresponding to the content selected from the human-animal language conversion TBL418, and memorize | stores it in RAM30 (step S506). Then, the key input process is terminated, and the flow returns to the flow of FIG. Returning to the flow of FIG. 12, the CPU 20 next executes an animal language output process.
[0107]
FIG. 20 is a diagram illustrating an example of a screen in the key input process according to the present embodiment. In the screen 7 for key input, the title display 7a indicates that the message is for the dog 2.
[0108]
A selection button 7b containing the text data 418b read from the human-animal language conversion TBL 418 is displayed. When all the selection buttons 7b cannot be displayed at a time, they are displayed in a scrollable manner. In addition, the selection button 7b that is currently selected is displayed in reverse video, for example.
[0109]
In addition, a selection button 7c and a cancel button 7d are displayed on the screen 7, and when the selection key 108b and the cancel key 108d are pressed, respectively, the screen 7 is highlighted and the user 4 is notified that the corresponding key has been input. Notify visually.
[0110]
In the flow of FIG. 12, for example, when the selection key 108b is pressed for a predetermined time for a relatively long time (step S122; YES), the CPU 20 executes a mode switching process (step S124).
[0111]
FIG. 17 is a flowchart for explaining the flow of the mode switching process in the present embodiment. As shown in the figure, the CPU 20 first refers to the screen frame data 414 to display a frame for mode switching on the display unit 52 (step S602).
[0112]
Next, when an operation for switching the high frequency mode is input (step S604; YES), the CPU 20 switches the high frequency mode flag 318 (step S606). When an operation for switching the sensation mode is input (step S608; YES), the CPU 20 switches the sensation mode flag 320 (step S610). If a predetermined end operation is input (step S612; YES), the mode switching process is ended, and the flow returns to the flow of FIG.
[0113]
FIG. 21 is a diagram showing an example of a screen in the mode switching process in the present embodiment. In the mode switching screen 7, the title display 8a indicates that the mode switching process is being executed. On the mode switching screen, a high frequency mode ON / OFF display 8b and a bodily sensation mode ON / OFF display 8c are displayed. The ON / OFF displays 8b and 8c are sequentially selected by the input of the up key 108a or the down key 108c. When the selection key 108b is input in the selected state, it means that the mode switching process has been input, and the CPU 20 switches between ON and OFF. When the cancel key 108d is pressed, the end operation of the mode switching process can be input.
[0114]
In the flow of FIG. 12, for example, when the cancel key 108d is pressed for a predetermined time for a relatively long time (step S126; YES), the CPU 20 executes a history display process (step S128).
[0115]
FIG. 22 is a diagram illustrating an example of a screen in the history display process according to the present embodiment. As shown in the figure, in the history display process, the history display unit 9a is displayed with reference to the history data 322. For example, a time 9b, an icon 9c indicating which voice of the dog 2 or the user 4 was used, and a content 9d are displayed. The icon 9c is displayed based on the voice identification flag 322b. The content 9d reads the text data 416b or 418b from the animal language human language conversion table 416 or the human language animal language conversion TBL 418 based on the voice identification flag 322b and the emotion identification code 322c, and displays the text.
[0116]
If the history display portion 9a cannot be displayed on the screen at a time, the history display portion 9a is displayed in a scrollable manner by the input of the up direction key 108a or the down direction key 108c. At this time, it is preferable to display with a bar 9e which time zone within one day (24 hours) the currently displayed history corresponds to.
[0117]
By viewing this history display, the user 4 can be used to know, for example, the personality and habit of the dog 2 and changes in physical condition.
[0118]
In FIG. 12, when no voice is input (step S102; NO), or when a voice is input but neither an animal voice nor a human voice is discriminated (step S114; NO), and a specific key operation is performed. Is not input (NO in step S118 → NO in S112 → NO in S126), the CPU 20 displays the clock screen 3 on the display unit 52 as shown in FIG. 23, for example (step S130).
[0119]
On the clock screen 3, for example, an analog clock 3a, a date 3b, and a day of the week 3c are displayed. Therefore, the user 4 can use the wristwatch type voice analysis apparatus 100 as a tool for communication with the dog 2 as well as a wristwatch.
[0120]
[Second Embodiment]
Next, a second embodiment of the speech analysis apparatus to which the present invention is applied will be described. Note that this embodiment can be realized basically by the same configuration as the first embodiment, and the same components are denoted by the same reference numerals and the description thereof will be omitted.
[0121]
FIG. 24 is a diagram showing an example of the appearance of the dog reader type speech analysis apparatus 200 in the present embodiment. As shown in the figure, the dog reader type speech analysis apparatus 200 includes a reader 202 that is used when the dog 2 is taken for a walk so that it can be pulled out / wound by a reel 204. At the tip of the reader 202, a collar 207 of the dog 2, a metal fitting 206 for attaching and detaching the reader 202, and the microphone unit 102 are provided. The user 4 grips the main body 208 or attaches it to a belt or the like with the clip unit 212 for use.
[0122]
The microphone unit 102 is connected to a control unit 120 and a power source built in the main body 208 by a signal line 210 disposed in the reader 202. By providing the microphone 102 at the tip of the reader 202, it is possible to collect sound more efficiently even under conditions where sound is likely to diffuse, such as outdoors.
[0123]
The dog reader type voice analysis apparatus 200 may be configured to transmit the voice signal collected by the microphone unit 102 via the wristwatch type voice analysis apparatus 100 worn by the user 4 and the data communication unit 114. In this case, the dog reader type voice analysis device 200 omits the voice analysis unit 12, the voice analysis ROM 16, the display unit 52, and the vibration unit 54, and uses the wristwatch type voice analysis device 100 worn by the user 4. It can be.
[0124]
Although the embodiments to which the present invention is applied have been described above, the application of the present invention is not limited to these embodiments, and appropriate modifications, additions, deletions, etc. of components are made without departing from the spirit of the invention. It doesn't matter.
[0125]
For example, the voice analysis device may be realized as a personal computer, a PDA (Personal Digital Assistant), or a multi-function phone.
[0126]
The voice analysis unit 12 may be configured by arithmetic processing of the CPU 20, and the voice analysis ROM 16 may be the same as the ROM 40. The key input unit 14 may be configured to have a touch panel on the display surface of the monitor unit 106.
[0127]
In the animal language output process, the synthesized speech data 418d is output when the speech data of the user 4 is output regardless of the collation result with the voiceprint data 410, and the speech data does not match the voiceprint data 410. A flow for additionally outputting data 418f may be used.
[0128]
【The invention's effect】
  According to the present invention, the following effects can be achieved.it can.
[0130]
  Claim1In the invention described inAccording to the userAre compared with a pre-registered voiceprint, and if they are different, the pre-registered human voice can be output in human language. Therefore, by outputting a human voice corresponding to the user's emotion to the animal, a training effect can be obtained that allows the animal to hear the human voice and become familiar with it. Further, if the voiceprint is the voiceprint of the person (for example, the owner) who is most familiar with the animal, the effect of calming the animal can be obtained even if the user is different from the person.
[0131]
  Claim2According to the invention described in (1), it is possible to carry and use the voice analysis device by making it attachable to the body. Therefore, it is not necessary to take out from the bag etc. one by one when using the speech analyzer, and the usability is improved.
[0132]
  Claim3, 4According to the invention described in the above, the history of the exchanges between the animals and the human is analyzed and used by storing the temporal history of the exchanges between the animals and the humans. be able to.
[0133]
  Claim5According to the invention described in, by generating a predetermined vibration corresponding to the analysis result of the voice of the animal and informing the user of the analysis result by bodily sensation, it is not necessary to read the human language text etc. Usability is improved and smoother communication becomes possible. Moreover, it becomes possible to know the analysis result even for a visually handicapped person or a hearing handicapped person.
[Brief description of the drawings]
FIG. 1 is a diagram showing an example of the appearance of a wristwatch type speech analysis apparatus according to a first embodiment.
FIG. 2 is a conceptual diagram showing an example of how to use a wristwatch type voice analysis apparatus.
FIG. 3 is a functional block diagram illustrating an example of a functional configuration.
FIG. 4 is a diagram showing an example of contents stored in a voice analysis ROM.
FIG. 5 is a diagram illustrating an example of a data configuration of an animal reference voice pattern.
FIG. 6 is a diagram illustrating an example of a data configuration of a human reference voice pattern.
FIG. 7 is a diagram showing an example of contents stored in a RAM.
FIG. 8 is a diagram showing an example of contents stored in a ROM.
FIG. 9 is a diagram showing an example of the data configuration of animal language / human language conversion TBL;
FIG. 10 is a diagram showing an example of the data structure of a human-animal language conversion TBL.
FIG. 11 is a diagram illustrating an example of a data configuration of a vibration pattern TBL.
FIG. 12 is a flowchart for explaining the main processing flow;
FIG. 13 is a flowchart for explaining the flow of voice analysis processing;
FIG. 14 is a flowchart for explaining the flow of human language output processing;
FIG. 15 is a flowchart for explaining the flow of animal language output processing;
FIG. 16 is a flowchart for explaining the flow of key input processing;
FIG. 17 is a flowchart for explaining the flow of mode switching processing;
FIG. 18 is a diagram showing an example of a screen in the human language output process.
FIG. 19 is a diagram showing an example of a screen in animal language output processing.
FIG. 20 is a diagram showing an example of a screen in key input processing.
FIG. 21 is a diagram showing an example of a screen in mode switching processing.
FIG. 22 is a diagram showing an example of a screen in history display processing.
FIG. 23 is a diagram showing an example of a clock display screen.
FIG. 24 is a diagram showing an example of the appearance of a dog reader type speech analyzer according to the second embodiment.
[Explanation of symbols]
2 dogs
4 users
10 Voice input part
12 Speech analysis unit
14 Key input section
16 ROM for voice analysis
162 Speech analysis program
164 Animal reference voice pattern
166 Human reference voice pattern
20 CPU
30 RAM
304 Animal attribute code
306 Human attribute code
310 Audio data
312 Voice input time data
314 Voice identification flag
316 Emotion identification code
318 High-frequency mode flag
320 Experience mode flag
322 history data
40 ROM
402 Animal Language Output Program
404 human language output program
406 Mode switching program
408 History output processing program
410 Voiceprint data
416 Animal language human language conversion TBL (table)
418 Human-animal language conversion TBL (table)
420 Vibration pattern TBL (table)
50 sound output section
52 Display
54 Excitation unit
60 Communication Department
100 Wristwatch type speech analyzer
102 Microphone part
104 Speaker section
106 Monitor unit
108 Key operation unit
110 Wristband
112 Vibrator
114 Data Communication Department
120 control unit
200 Dog Reader Type Voice Analyzer
202 reader
204 reel
206 metal fittings
208 body
210 Signal line
212 Clip part

Claims (5)

In the voice analysis device provided with the voice analysis device body,
First voice input means for inputting animal voice;
First voice analysis means for analyzing the voice input by the first voice input means;
First output means for displaying and outputting characters or images corresponding to the analysis result by the first voice analysis means as human language;
After the human language is displayed and output by the first output means, a second voice input means for inputting the user's voice in order to respond to the displayed and output content;
Second voice analysis means for analyzing the voice input by the second voice input means;
Second output means for outputting voice corresponding to an analysis result by the second voice analysis means as an animal language understandable by the animal;
A first registration means for registering a user's voiceprint;
A second registration means for registering speech in human language having a predetermined meaning;
Determination means for determining whether or not the voice input by the second voice input means matches the voiceprint registered by the first registration means;
A fourth output means for outputting the speech registered by the second registration means in human language when it is determined by the determination means that they do not match;
A speech analysis apparatus comprising: The speech analysis apparatus according to claim 1 , further comprising mounting means for mounting the speech analysis apparatus main body on a user's body. 3. The storage device according to claim 1, further comprising a storage unit that stores the voice input by the first voice input unit and the time when the voice is input by the first voice input unit in association with each other. The speech analyzer described. The speech analysis apparatus according to claim 3 , further comprising a fifth output unit that outputs the speech stored in the storage unit in association with the time. Sound analysis apparatus according to any one of claims 1-4, characterized in that it comprises vibration means for generating a predetermined vibration corresponding to the analysis result by said first sound analysis unit.

Images