KR20180124226A - WEARABLE DEVICE, IoT DEVICE AND VEHICLE VOICE RECOGNITION OPERATING SYSTEM AND METHOD - Google Patents

Abstract

The present invention relates to a wearable device, an IoT device, and a vehicle voice recognition operating system and a method thereof. The system includes: a voice recognition device for recognizing a voice command of a user and converting the voice command into a digital signal by dictating the voice command to generate dictation data; and a command execution device connected to at least one of the wearable device, the internet of things (IoT) device, a portable terminal, a portable storage device, a navigation device, and an AVN of a vehicle, and performing a command using the dictation data. According to the present invention, the accuracy of voice command recognition intended by the user can be increased by firstly grasping the intention of the user included in the voice command through similar numeric data and correcting an error of the voice command through examination data.

Description

FIELD OF THE INVENTION [0001] The present invention relates to a wearable device, an IoT device, and a vehicle voice recognition operating system and method.

More particularly, the present invention relates to a wearable device, an IoT (Internet of Things) device, and a wearable device installed in a vehicle for recognizing a voice command of a user to perform a command, an IoT device, System, and a wearable device, an IoT device, and a vehicle voice recognition operating method using the same.

Speech Recognition analyzes the user's voice input through a microphone and extracts the features to recognize the results of the words or sentences close to the previously entered words as an instruction word and performs the operation corresponding to the recognized instruction word Technology.

When the speech recognition technology is applied to a wearable device, an Internet of Things (IoT) device, and a vehicle, a user can obtain desired information only by voice without operating the device directly using his or her hand, .

However, the conventional wearable device, IoT device, and voice recognition technology applied to the vehicle have a capability of recognizing each of the various search domains such as destination, address, and music by performing speech recognition using only phoneme array probability information without knowing domain information .

In addition, since the recognition result is confirmed only by the phoneme information without semantic information, an error occurs in the semantic classification of the grammatical elements such as the investigation, resulting in semantic confusion.

Korean Patent Application No. 10-2000-0037217 (Jun. 30, 2000) " Speech Recognition-Based Operational Toys Set and Control Method Thereof " Korean Patent Laid-Open No. 10-2008-0103162 (2008.11.27) " Vehicle Speech Recognition System with Microphone Array "

An object of the present invention is to provide a wearable device, an Internet of Things (IoT) device, and a vehicle voice recognition operating system that recognize a user's voice and perform a user's command.

It is also an object of the present invention to provide a vehicle voice recognition operating method capable of performing an instruction using the above wearable device, IoT device and vehicle voice recognition operating system.

The wearable device, the IoT device, and the vehicle voice recognition operating system according to one aspect for realizing the object of the present invention recognize a voice command of a user and convert the voice command into a digital signal by dictation A speech recognition device for generating dictation data; And an instruction execution device connected to at least one of a wearable device, an IoT device, a portable terminal, a mobile storage device, a navigation device, and an AVN of a vehicle, and performing an instruction using the dictation data.

The speech recognition apparatus includes: a speech recognition unit for recognizing the speech command and generating speech data; A noise removing unit for removing noise of the voice data; And an A / D conversion unit for dictating the voice data from which the noise is removed, and converting the voice data into a digital signal to generate dictation data.

Here, the command execution device may include: an object name data storage unit that includes a plurality of object name data indicating object names, and categorizes and stores a plurality of object name data; A domain data storage unit for storing a plurality of domain data including a plurality of utterance patterns which can be ignited for each domain and a plurality of commands interlocked with each of the plurality of utterance patterns; An object name extracting unit for analyzing the dictation data and extracting an object name included in the dictation data to generate name data; And a command extracting unit for analyzing the utterance pattern of the dictation data using the plurality of domain data, extracting command information and domain information interlocked with the utterance pattern, and generating command data including the extracted domain information and command information .

In addition, the command execution device may compare the plurality of object name data using the name data and the command data to calculate respective similarity values, and compare the object name data having the highest similarity value with the calculated similarity value, A similarity comparison unit for generating similarity numerical data including category information including object name data; Comparing the plurality of pieces of object name data included in the category information of the similar numeric data with the name data to calculate respective similarity values and comparing the object name data having the highest similarity value with the calculated similarity value, A data review unit for generating the data; An instruction execution unit for comparing the similar numeric data with the review data to select data having a higher similarity value, and to perform an instruction using the selected data and the instruction data; A communication unit connected to at least one of the wearable device, the IoT device, the portable terminal, the portable storage device, the navigation device, and the AVN of the vehicle by wireless or wire communication; And a controller for controlling the flow and overall operation of the data.

In addition, the instruction execution device may further include a data correction unit that corrects the dictation data using the review data if the similar numeric data and the review data are the review data.

According to another aspect of the present invention, there is provided a wearable device, an IoT device, and a vehicle voice recognition operating method, wherein the voice recognition device recognizes a voice command of a user and generates dictation data step; And an instruction execution device performing an instruction using the dictation data.

The step of generating the dictation data may include: generating the voice data by recognizing a voice command of the user; Removing noise of the voice data; Dictating the voice data and converting the voice data into a digital signal to generate dictation data; And transmitting the dictation data to the command execution device.

The step of executing the command may further include: receiving the dictation data; The object name extracting unit analyzes the dictation data and extracts the object name included in the dictation data to generate name data; Analyzing an utterance pattern of the dictation data using a plurality of the domain data, extracting command information and domain information interlocked with the utterance pattern, and generating command data including extracted domain information and command information ; And the similarity comparison unit compares the plurality of object name data using the name data and the command data to calculate respective similarity values, and the object name data having the highest similarity value, the calculated similarity value, and the object name data And generating similar numeric data including the included category information.

The step of executing the command may further include comparing the plurality of object name data and the name data in which the data review unit is included in the category information of the pseudo-numeric data to calculate each similarity value, Generating review data including object name data and the calculated similarity value; And the command execution unit may compare the similar numeric data and the review data to select data having a higher similarity value, and executing the command using the selected data and the command data.

The step of generating the pseudo-numeric data may include: calculating a plurality of similarity values by comparing a plurality of object name data included in the object name data storage unit using the name data and the command data; And generating the similar numeric data including the object name data having the highest similarity value among the object name data, the calculated similarity value, and the category information including the object name data.

The step of generating the review data may include the steps of: calculating a similarity value by comparing the plurality of object name data included in the category information of the similar numeric data with the name data; And generating the review data including the object name data having the highest similarity value among the object name data and the calculated similarity value.

The step of executing the command may further include comparing the similar-value data with a similarity value included in the review data; Selecting data having a similarity value higher than the similarity value data and the review data; Determining whether an instruction can be executed using the object name data included in the command data and the data having a higher similarity value; And performing the instructions.

The step of executing the instruction may further include a step of, when the instruction execution unit selects the similar numeric data and the review data, if the data selected by the instruction execution unit is the review data, And then modifying the dictation data.

According to the wearable device, the IoT device, and the vehicle voice recognition operating system and method according to the embodiment of the present invention,

First, through the similar numerical data, the intention of the user included in the voice command can be grasped first and the error of the voice command can be corrected through the review data, so that the accuracy of the voice command recognition that the user intends can be improved.

Second, the domain information is analyzed by analyzing the voice command, and the name data and the object name data are compared using the voice command, thereby improving the accuracy of selecting the data intended by the user among the object name data having the same sound.

Third, the accuracy of performing a voice command can be improved by calculating the similarity value with object name data.

Fourth, the speech pattern of the voice command can be grasped by using a plurality of domain data, and the accuracy of grasping the intention of the user can be improved.

Fifth, the voice command error of the user can be corrected through the review data, thereby improving the accuracy of command execution.

1 is a conceptual view schematically showing a wearable device, an IoT device, and a vehicle voice recognition operating system according to an embodiment of the present invention.
FIG. 2 is a conceptual schematic diagram of the speech recognition apparatus shown in FIG. 1. FIG.
3 is a schematic conceptual diagram of the instruction execution apparatus shown in FIG.
4 is a flowchart illustrating a wearable device, an IoT device, and a vehicle voice recognition method according to an embodiment of the present invention.
FIG. 5 is a flowchart for explaining a process of generating the voice data shown in FIG.
FIG. 6 is a flowchart for explaining a process of executing the command shown in FIG.
FIG. 7 is a flowchart for explaining a process of generating the pseudo-numerical data shown in FIG.
FIG. 8 is a flowchart for explaining the process of generating the review data shown in FIG.
FIG. 9 is a flowchart for explaining a process of executing the command shown in FIG.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. Note that, in the drawings, the same components are denoted by the same reference symbols as possible. Further, the detailed description of known functions and configurations that may obscure the gist of the present invention will be omitted. For the same reason, some of the components in the drawings are exaggerated, omitted, or schematically illustrated.

FIG. 1 is a conceptual view that schematically shows a wearable device, an IoT device, and a vehicle voice recognition operating system according to an embodiment of the present invention. FIG. 2 is a schematic diagram of a speech recognition device shown in FIG. 1, 1 is a schematic conceptual diagram of an instruction execution device shown in FIG.

1 and 3, the wearable device, the Internet of Things (IoT) device, and the vehicle voice recognition operating system 100 of the present invention include a voice recognition device 110 and an instruction execution device 120.

The speech recognition apparatus 110 recognizes the voice command of the user, converts the voice command into a digital signal by dictation, generates the dictation data, and transmits the dictation data to the command execution apparatus 120 do.

Here, the speech recognition apparatus 110 may be provided as an embedded device or a server.

The speech recognition apparatus 110 includes a speech recognition unit 111, a noise removal unit 112, an A / D conversion unit 113, a communication unit 114, and a control unit 115.

The speech recognition unit 111 recognizes the voice command of the user to generate voice data, and the noise removing unit 112 removes noise of the voice data.

The A / D conversion unit 113 then dictates the voice data from which the noise has been removed, and converts the voice data into a digital signal to generate dictation data.

Here, the dictation data is data that recognizes a voice command of the user and is expressed in characters.

The control unit 115 controls the overall operation of the voice recognition apparatus 110. The communication unit 114 is connected to the command execution unit 120, which will be described later, in a wireless or wired manner.

The instruction execution device 120 is connected to at least one of a wearable device, an IoT device, a portable terminal, a portable storage device, a navigation device, and an AVN of a vehicle, receives dictation data from the voice recognition device 110, To execute the command.

The command execution unit 120 includes an entity name data storage unit 121, a domain data storage unit 122, an entity name extraction unit 123, an instruction extraction unit 124, a similarity comparison unit 125, An instruction execution unit 127, a data correction unit 128, a communication unit 129, and a control unit 131. [0033]

The entity name data storage unit 121 includes a plurality of entity name data indicating entity names, and stores a plurality of entity name data into categories.

The domain data storage unit 122 stores a plurality of domain data including a plurality of utterance patterns that can be ignited for each domain and a plurality of commands interlocked with each of the plurality of utterance patterns.

Here, the domain data is divided into search domains such as place, address, music, and the like, and includes a plurality of commands for each domain, and includes an utterance pattern interlocked for each command.

For example, there is POI (Point of Interest) domain data indicating a place search among domain data, and a command "SearchPlace" is included in the commands included in the domain. This command is called "Find <POI> Pattern.

The object name extracting unit 123 analyzes the dictation data and extracts the object name included in the dictation data to generate name data.

Here, the name data may be a noun included in the voice command of the user, for example, a song title, a place name, or a noun such as a restaurant name.

The command extracting unit 124 analyzes the utterance pattern of the dictation data using a plurality of domain data, extracts command information and domain information interlocked with the utterance pattern, generates command data including the extracted domain information and command information do.

For example, when the user gives a voice command "Find Post Office", the command extracting unit 126 can extract the command information "SearchPlace" and the domain information "POI" by analyzing the utterance pattern.

The similarity comparison unit 125 compares a plurality of object name data using name data and command data, and calculates each similarity value. The similarity comparison unit 125 compares the object name data having the highest similarity shake with the calculated similarity value and object name data And generates similar numeric data including category information.

Here, the similarity comparison unit 125 may use the domain information included in the command data and the phoneme arrangement of the name data to calculate the similarity value differently according to the domain information, even though the data indicates the same phoneme arrangement.

For example, if there are two pieces of data having the same phoneme array of "Post Office" among the object name data, and the respective places and the songs are represented, data indicating the place rather than the data representing the song through the domain information included in the command data, The figures may be higher.

The data review unit 126 compares the plurality of pieces of object name data included in the category information of the name data and the similar numeric data to calculate respective similarity values, and includes the object name data having the highest similarity value and the calculated similarity value And generates review data.

Here, the data review unit 126 compares the phoneme arrays of the name data with the object name data included in the category corresponding to the command word of the command data, and generates the review data.

The instruction execution unit 127 compares the similar numeric data with the review data, selects data having a higher similarity value, and executes the instruction using the selected data and the instruction data.

Here, if the similar numeric data and the review data include the same data, one can arbitrarily select one to execute the instruction.

The data correction unit 128 corrects the dictation data using the review data if the data selected by the command execution unit 127 among the similar numerical data and review data is review data.

Accordingly, even if a similar entity name is spoken, the user can not correct the entity name included in the voice command of the user, so that the user can correct the entity name with the intended entity name, thereby improving the accuracy of executing the voice command of the user.

The communication unit 129 is connected to at least one of the wearable device 10, the IOT device 20, the portable terminal 30, the portable storage device 40, the navigation device 50 and the AVN 60 of the vehicle, And the control unit 131 controls the flow of data and overall operation.

FIG. 4 is a flowchart for explaining a wearable device, an IoT device, and a vehicle voice recognition method according to an embodiment of the present invention. FIG. 5 is a flowchart illustrating a process of generating voice data shown in FIG. FIG. 6 is a flowchart for explaining a process of executing the command shown in FIG. 4, and FIG. 7 is a flowchart for explaining a process of generating the pseudo-numerical data shown in FIG.

FIG. 8 is a flowchart for explaining a process of generating review data shown in FIG. 6, and FIG. 9 is a flowchart for explaining a process of executing the instructions shown in FIG.

1 and 4, the wearable device, the IoT device, and the vehicle voice recognition operating method of the present invention recognize voice commands of a user by voice recognition device 110 to generate dictation data (step S110 ).

Then, the instruction execution device 120 executes the instruction using the dictation data (step S120).

The process of generating the dictation data by the speech recognition apparatus 110 (step S110) will be described in detail with reference to FIGS. 2 and 5. FIG.

The voice recognition unit 111 of the voice recognition apparatus 110 recognizes the voice command of the user and generates voice data (step S111).

Then, the noise removing unit 112 removes the noise of the audio data (step S112).

Thereafter, the A / D converter 113 dictates the voice data from which noise has been removed, converts the voice data into a digital signal, and generates dictation data (step S113).

Then, the dictation data is transmitted to the command execution device 120 (step S114).

1, 3, and 6, a description will be made in detail of a process (step S120) in which the command execution unit 120 executes an instruction.

First, the dictation data is received from the speech recognition apparatus 110 (step S121), the entity name extraction unit 123 analyzes the dictation data, extracts the entity name included in the dictation data, and generates name data (step S122) .

The command extracting unit 124 analyzes the utterance pattern of the dictation data using a plurality of domain data, extracts the command information and domain information interlocked with the utterance pattern, extracts the extracted domain information and command data (Step S123).

Thereafter, the similarity comparison unit 125 compares the plurality of object name data using the name data and the command data, and calculates each similarity value. Then, the object name data having the highest similarity value, the calculated similarity value and the object name data And generates similar numeric data including the included category information (step S124).

Then, the data review unit 126 compares the plurality of pieces of object name data and name data included in the category information of the similar-value data to calculate respective similarity values, and the object name data having the highest similarity value and the calculated similarity value And generates review data (step S125).

Thereafter, the command execution unit 127 compares the similar numeric data with the review data, selects data including a higher similarity value, and executes the command using the selected data and the command data (step S126).

Referring to FIGS. 3 and 7, the process of generating similar-number data at step S120 (step S124) will be described in detail.

The similarity comparison unit 125 compares a plurality of object name data included in the object name data storage unit 123 using the name data and the command data, and calculates each similarity value (step S124a).

Then, pseudo-numeric data including object name data having the highest similarity value among the object name data and category information including the calculated similarity value and object name data is generated (step S124b).

The process of generating review data in step S120 (step S125) will be described in detail with reference to FIGS. 3 and 8. FIG.

The data review unit 126 compares the plurality of pieces of object name data included in the category information of the similar numeric data with the name data and calculates each similarity value (step S125a).

Then, review data including the object name data having the highest similarity value among the object name data and the calculated similarity value is generated (step S125b).

The process of executing the command of step S120 (step S126) will be described in detail with reference to FIG. 3 and FIG.

The command execution unit 127 compares the similar-value data and the similarity value included in the review data (step S126a).

Then, the command execution unit 127 selects data having a similarity value among the similar-value data and the review data (step S126b).

If the data selected in step S126b is the review data, the data correction unit 128 modifies the dictation data using the review data (step S126c) and does not modify the dictation data if it is similar-value data.

Then, in step S126b, it is determined whether the command can be executed by using the object name data and the command data included in the selected data (step S126d).

In step S126d, a command is sent to at least one of the wearable device 10, the IOT device 20, the portable terminal 30, the portable storage device 40, the navigation device 50, and the AVN 60 of the vehicle If there is no entity name to perform, the user can request the voice command again.

The IoT device 20, the portable terminal 30, the portable storage device 40, the navigation device 40, and the navigation device 40. In this case, the object name data is a song title and the command data is a command to reproduce a song. If the song title indicated by the entity name data does not exist in at least one of the AVN 50 and the AVN 60 of the vehicle, the user can request the voice command again.

On the other hand, in step S126d, at least one of the wearable device 10, the IOT device 20, the portable terminal 30, the portable storage device 40, the navigation device 50 and the AVN 60 of the vehicle If there is an entity name matching the entity name for executing the command, this is done using the command data (step S126e).

[0051] The operation of the vehicle voice recognition operating system 100 of the present invention will be described with reference to the above-described configurations and steps. For example, when the user gives a voice command " Play Paris " , The voice recognition unit 111 of the voice recognition apparatus 110 recognizes it and generates voice data, and the noise removing unit 112 removes the noise of the voice data.

The A / D converter 113 then dictates the noise-removed speech data, converts the resulting speech data into a digital signal to generate dictation data, and transmits the generated digital data to the command execution unit 120.

Then, the command execution unit 120 extracts the object name included in the dictation data by analyzing the dictation data by the object name extracting unit 123 to generate name data.

Here, the name data represents the " Paris " part of the voice command of the user.

Thereafter, the command extracting unit 124 analyzes the utterance pattern of the dictation data using a plurality of domain data stored in the domain data storage unit 122, and the command information " Play " associated with the utterance pattern and the command information And extracts the domain information " VAE " to generate command data including the extracted domain information and command information.

Then, the similarity comparison unit 125 compares a plurality of object name data stored in all categories of the object name data storage unit 121 by using the name data and the command data, calculates each similarity value, Similarity numerical data including object name data and category information including the calculated similarity value and object name data is generated.

Assuming that the object name data 1 among the object name data of all categories of the object name data storage unit 121 gives the song "Paris" and the object name data 2 the homonym "Paris"

The similarity value is the phoneme array of the name data " Paris " and the similarity value between the name data and the object name data 1 through the command information " Play " of the command data is 74 and the similarity value with the object name data 2 is 66, The similarity value can be calculated differently from the command information.

The similar numeric data is generated including category information "Music" including the entity name data 1, the calculated similarity value 74, and the entity name data 1.

Thereafter, the data review unit 126 compares the plurality of pieces of object name data included in the category information " Music " of the similar numeric data with the name data to calculate respective similarity values, and the object name data having the highest similarity value And generates review data including the calculated similarity value.

Assuming that the object name data 3 among the object name data included in the category "Music" of the category of the object name data storage unit 123 indicates the song "Paris" and the object name data 4 indicates the song "Paradise"

When the similarity value between the name data and the object name data 3 is 72 and the similarity value with the object name data 4 is 70 by the data review unit 126, the review data includes the object name data 3 and the calculated similarity value 72 .

Here, the similarity value between the name data, the object name data 3 and the object name data 4 is calculated by comparing only the phoneme arrangement.

Then, the instruction execution unit 127 compares the similarity value 74 included in the similarity numerical data with the similarity value 72 included in the review data, and selects similar numerical data having a higher similarity value.

Then, the wearable device 10, the IOT device 20, the portable terminal 30, and the portable device 30 are accessed using the "Paris" object name data 1 included in the similar numeric data and the command information "Play" And searches for a reproduction file corresponding to " Paris " on at least one of the storage devices 40 to reproduce music.

Here, if the similar-value data and the review data show the same similarity value, one can arbitrarily select one to carry out the instruction.

As another example, if the similarity value of the object name data 4 is 77 in the data review unit 126 and the review data is generated including the object name data 4 and the similarity value 77, the command execution unit 127 selects the review data , The data correction unit 128 modifies the dictation data to "Play Paradise" by using "Paradise" having the object name data 4, and then the command execution unit 127 executes the command.

For example, the operation of the wearable device, the IoT device, and the vehicle voice recognition operating system 100 of the present invention has been described with a simple example in order to facilitate a quick understanding.

It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined in the appended claims. It will be possible.

110 ... voice recognition unit 111 ... voice recognition unit
112 ... Noise removing unit 113 ... A / D converting unit
120 ... command execution unit 121 ... object name data storage unit
122 ... domain data storage unit 123 ... object name extracting unit
124 ... command extracting unit 125 ... similarity comparing unit
126 ... data review unit 127 ... command execution unit
128 ... data correction unit 129 ... communication unit
131 ... control unit

Claims (13)

A speech recognition device for recognizing a voice command of a user, converting the voice command into a digital signal and generating dictation data; And
A wearable device, an instruction execution device connected to at least one of a wearable device, an Internet of Things (IoT) device, a portable terminal, a portable storage device, a navigation device, and an AVN of a vehicle and performing an instruction using the dictation data Device, IoT device and vehicle voice recognition operating system. The method according to claim 1,
The speech recognition apparatus includes:
A voice recognition unit for recognizing the voice command and generating voice data;
A noise removing unit for removing noise of the voice data; And
And an A / D converter for converting the voice data into a digital signal and generating dictation data, wherein the voice data is obtained by removing the noise from the voice data. 3. The method of claim 2,
The instruction execution device comprises:
An object name data storage unit for storing a plurality of object name data indicating object names and categorizing and storing the plurality of object name data;
A domain data storage unit for storing a plurality of domain data including a plurality of utterance patterns which can be ignited for each domain and a plurality of commands interlocked with each of the plurality of utterance patterns;
An object name extracting unit for analyzing the dictation data and extracting an object name included in the dictation data to generate name data; And
A command extracting unit for analyzing a speech pattern of the dictation data using a plurality of the domain data, extracting command information and domain information interlocked with the utterance pattern, and generating command data including extracted domain information and command information The IoT device, and the vehicle voice recognition operating system. The method of claim 3,
The instruction execution device comprises:
Comparing the plurality of object name data using the name data and the command data to calculate respective similarity values, and comparing the object name data having the highest similarity value with the category information including the calculated similarity value and the object name data A similarity comparison unit for generating similarity value data including the similarity data;
Comparing the plurality of pieces of object name data included in the category information of the similar numeric data with the name data to calculate respective similarity values and comparing the object name data having the highest similarity value with the calculated similarity value, A data review unit for generating the data;
An instruction execution unit for comparing the similar numeric data with the review data to select data having a higher similarity value, and to perform an instruction using the selected data and the instruction data;
A communication unit connected to at least one of the wearable device, the IoT device, the portable terminal, the portable storage device, the navigation device, and the AVN of the vehicle by wireless or wire communication; And
A controller for controlling the flow and overall operation of the data, the IoT device and the vehicle speech recognition operating system. 5. The method of claim 4,
The instruction execution device comprises:
Further comprising a data correcting unit for correcting the dictation data using the review data if the similar numeric data and the review data are the review data. The speech recognition apparatus recognizing a voice command of the user to generate dictation data; And
Wherein the command execution device executes an instruction using the dictation data, the IoT device, and the vehicle voice recognition operating method. The method according to claim 6,
Wherein the generating the dictation data comprises:
Recognizing a voice command of the user and generating the voice data;
Removing noise of the voice data;
Dictating the voice data and converting the voice data into a digital signal to generate dictation data; And
And transmitting the dictation data to the command executing device. The wearable device, the IoT device, and the vehicle voice recognition operating method according to claim 1, 8. The method of claim 7,
The step of performing the instruction comprises:
Receiving the dictation data;
The object name extracting unit analyzes the dictation data and extracts the object name included in the dictation data to generate name data;
Analyzing an utterance pattern of the dictation data using a plurality of the domain data, extracting command information and domain information interlocked with the utterance pattern, and generating command data including extracted domain information and command information ; And
The similarity comparison unit compares the plurality of object name data using the name data and the command data to calculate respective similarity values, and the object name data having the highest similarity value, the calculated similarity value, and the object name data are included And generating pseudo-numerical data that includes the category information. The wearable device, the IoT device, and the vehicle voice recognition operating method according to claim 1, 9. The method of claim 8,
The step of performing the instruction comprises:
The data review unit compares the plurality of pieces of object name data included in the category information of the similar numeric data with the name data to calculate each similarity value and includes the object name data having the highest similarity value and the calculated similarity value Generating review data; And
Wherein the instruction execution unit further comprises a step of comparing the similar numeric data with the review data to select data having a higher similarity value and executing the instruction using the selected data and the instruction data. And vehicle voice recognition. 10. The method of claim 9,
Wherein the step of generating the similar-
Comparing the plurality of object name data included in the object name data storage unit using the name data and the command data to calculate respective similarity values; And
And generating the similar numerical data including object name data having a highest similarity value among the object name data, category similarity value, and category information including the object name data. How IoT devices and vehicle voice recognition work. 11. The method of claim 10,
Wherein the step of generating the review data comprises:
Comparing the plurality of pieces of entity name data included in the category information of the pseudo-numeric data with the name data to calculate respective similarity values; And
And generating the review data including the object name data having the highest similarity value among the object name data and the calculated similarity value. 12. The method of claim 11,
The step of performing the instruction comprises:
Comparing the similar-value data with a similarity value included in the review data;
Selecting data having a similarity value higher than the similarity value data and the review data;
Determining whether an instruction can be executed using the object name data included in the command data and the data having a higher similarity value; And
Wherein the IoT device and the vehicle voice recognition operating method comprise the steps of: 13. The method of claim 12,
The step of performing the instruction comprises:
If the data selected by the instruction execution unit is the review data,
After the instruction execution unit selects the similar numeric data and the review data,
And the data correction unit corrects the dictation data using the review data. &Lt; RTI ID = 0.0 &gt; 11. &lt; / RTI &gt;

Images