Electronic device and method of controlling voice recognition thereof

Abstract

An electronic device capable of controlling voice recognition is disclosed. When an instruction included in a user's utterance exists in a database, the electronic device increases a score of a category corresponding to a word included in the user's utterance in the database. When the instruction does not exist in the database, the electronic device checks whether the score of the category corresponding to the word is equal to or greater than a preset value. As a result of the check, when the score is equal to or greater than the preset value, the electronic device registers the instruction in the database so that the instruction is included in the category corresponding to the word.

Description

Technical Field

[0001] This disclosure relates to an electronic device and a method for controlling speech recognition. Background Technology

[0002] With the development of speech recognition technology, it has been applied to various electronic devices, such as TVs, speakers, and mobile phones. A typical method for controlling speech recognition converts and recognizes the user's speech as data, analyzes it, and executes the content of the recognized speech. At this point, the meaning of the recognized speech is checked by querying a pre-built database.

[0003] For example, when a user says "Let me know the weather tomorrow," the user's instruction is "Let me know," and the target of the instruction is "tomorrow's weather." A category called "weather" exists in the database. This category includes several words, such as "tomorrow's weather," "today's weather," "New York weather," "Washington weather," "weather forecast," and "weather conditions." Furthermore, various words representing the user's desired instruction regarding the "weather" category (such as "Let me know," "Tell me," and "How is it?") are pre-registered in the database as weather-related instructions to inform the user. Therefore, when a user says "Let me know the weather tomorrow," the speech recognition system interprets the utterance as meaning to provide information about tomorrow's weather and performs the corresponding action.

[0004] However, traditional speech recognition systems rely on pre-built databases to understand user intent. Therefore, speech recognition fails when undefined words or commands are input. For example, if a user says, "Do you want to know tomorrow's weather?", and the command "Do you want to know...?" is not pre-registered in the database, the speech recognition system may not recognize what action to take in response to "tomorrow's weather." In this case, the system determines that recognition has failed and does not execute the user's intended action. Therefore, when a user wants to execute a command, they can only do so if they speak a command pre-defined in the database. Summary of the Invention

[0005] Technical issues

[0006] One or more embodiments of this disclosure provide a method for understanding a user's intent and performing corresponding operations even when the instructions in the user's utterance are not predefined in a database.

[0007] One or more embodiments of this disclosure provide a method that increases the success probability of speech recognition by proactively incorporating unexpected instructions spoken by various users into new instructions later, even when no instructions in a user’s speech have been previously defined in a database.

[0008] Technical solution

[0009] According to one aspect of the embodiments, an electronic device including a processor is provided.

[0010] The processor can be configured to: based on the existence of an instruction included in a user's utterance in a preset database, increase the score of the category corresponding to the word included in the user's utterance in the database. The processor can also be configured to: based on the absence of the instruction in the database, check whether the score of the category corresponding to the word is equal to or greater than a preset value, and based on the score being equal to or greater than the preset value, register the instruction in the database, such that the instruction is included in the category corresponding to the word.

[0011] The processor may be configured to: examine the category for each of the words included in the discourse, and identify the category corresponding to the instruction in the discourse based on the score of the examined category.

[0012] The processor can be configured to: create a new category in the database for the words included in the utterance, based on the fact that the words included in the utterance do not exist in the database.

[0013] The processor may be configured to: select one of a plurality of instructions included in the category for the word as the instruction for the word, based on the fact that the instruction in the utterance does not exist in the database.

[0014] The processor can be configured to select any one of a plurality of categories for each of the plurality of words based on a score, based on the plurality of words included in the discourse.

[0015] The processor can be configured to receive checks from the user for selected instructions.

[0016] An electronic device according to this disclosure may include a memory, wherein the memory is configured to store the database including fields for each of the words, the instructions, the categories, and the scores.

[0017] An electronic device according to this disclosure may include a communication interface, wherein the communication interface is configured to perform communication for querying the database on an external database server storing the database.

[0018] An electronic device according to this disclosure may include an actuator, wherein the actuator is configured to perform an operation corresponding to the word and the instruction.

[0019] An electronic device according to this disclosure may include an output interface, wherein the output interface is configured to output the result of the operation performed by the actuator.

[0020] An electronic device according to this disclosure may include an interface configured to send the result of an operation performed by the actuator to an external output device, such that the external output device displays the result of the operation.

[0021] An electronic device according to this disclosure may include a microphone, wherein the microphone is configured to receive the sound of the speech and send the received sound to a recognizer.

[0022] An electronic device according to this disclosure may include a receiver, wherein the receiver is configured to receive a signal for the sound from an external device including a microphone for receiving the sound of the speech, and to send the received signal to the recognizer.

[0023] According to one aspect of the embodiments, a method for controlling speech recognition of an electronic device is provided, wherein the method includes: a) increasing the score of a category in the preset database corresponding to a word in the user's speech based on the existence of an instruction included in the user's speech in the preset database; and b) checking whether the score of the category corresponding to the word is equal to or greater than the preset value based on the absence of the instruction in the database, and registering the instruction in the database based on the equality or difference of the score with respect to the preset value, such that the instruction is included in the category corresponding to the word.

[0024] In step b), based on the inclusion of multiple words in the discourse, the category including the instruction can be identified based on the score of the category for each of the multiple words.

[0025] The speech recognition method for controlling electronic devices according to this disclosure may further include: c) creating a new category for the word in the database based on the fact that the word included in the utterance does not exist in the database.

[0026] The speech recognition method for controlling an electronic device according to this disclosure may further include: d) selecting any one of the instructions in the category for the word, based on the fact that the instruction does not exist in the database.

[0027] In step d), based on the plurality of words being included in the discourse, any one of the categories for each of the words can be identified based on the score.

[0028] The voice recognition method for controlling electronic devices according to this disclosure may further include: e) receiving a check from a user for an instruction selected in step d).

[0029] The speech recognition method for controlling an electronic device according to this disclosure may further include: f) performing an operation on the words in the speech according to the instruction selected in step d).

[0030] Beneficial effects

[0031] According to this disclosure, if a user utters a word from a category after frequently using that category, the instruction in the utterance is a previously unregistered instruction, and thus the instruction is registered as a new instruction in the corresponding category. Therefore, speech recognition is appropriately performed to suit the intentions of various users, and the number of instances identified as failed utterances is reduced.

[0032] Furthermore, even if the instruction included in the current input utterance does not exist in the category database, the instruction that matches the user's intent is selected. Therefore, even if the instruction is classified as a failed utterance, the instruction assumed to be the user's intent is selected and executed, without immediately obtaining a result indicating recognition failure. Attached Figure Description

[0033] The above and other aspects, features and advantages of certain embodiments of this disclosure will become more apparent from the following description taken in conjunction with the accompanying drawings.

[0034] Figure 1 It is a block diagram of an electronic device according to this disclosure.

[0035] Figure 2 It is shown Figure 1 A schematic diagram of the database structure.

[0036] Figure 3 and Figure 4 It is by Figure 1 A flowchart of a method for controlling speech recognition executed by an electronic device.

[0037] Figures 5 to 8 They are shown respectively Figure 1 The modified form is shown in the diagram. Detailed Implementation

[0038] In the following, various embodiments of the present disclosure will be described in more detail with reference to the accompanying drawings.

[0039] However, the technical concept, core configuration, and operation of this disclosure are not limited to the configurations or operations described in the following examples. In describing this disclosure, detailed descriptions of known technologies or configurations related to this disclosure will be omitted if it is determined that such detailed descriptions would unnecessarily obscure the subject matter of this disclosure.

[0040] In this disclosure, the expressions “having,” “may have,” “include,” “may include,” etc., indicate the presence of a corresponding feature (e.g., numerical value, function, operation, component such as a part), and do not exclude the presence of other features.

[0041] In this disclosure, expressions such as “A or B”, “at least one of A and / or B”, “one or more of A and / or B” can include all possible combinations of the items listed together. For example, “A or B”, “at least one of A and B” or “at least one of A or B” can indicate all of the following: (1) the case that includes at least one A, (2) the case that includes at least one B, or (3) the case that includes both at least one A and at least one B.

[0042] In the description of this disclosure, unless the shape, structure, position, etc. are necessary for the implementation of this disclosure, the shape, structure, position, etc. do not limit the scope of protection of this disclosure.

[0043] The expression “configured (or set) as” as used in this disclosure may be replaced by expressions such as “suitable for,” “capable of,” “designed to,” “suitable for,” “manufactured as,” or “capable”, depending on the context. The term “configured (or set) as” may not necessarily mean “specifically designed for” in hardware. Rather, the expression “device configured as” may mean that the device can “do” together with other devices or components. For example, “configured (or set) as subprocessors to perform A, B, and C” may mean a dedicated processor (e.g., an embedded processor) for performing the respective operations or a general-purpose processor (e.g., a central processing unit (CPU) or application processor) that can perform the respective operations by executing one or more software programs stored in a memory device.

[0044] In the description of this disclosure, the term "electronic device" can refer to all types of electronic devices, such as televisions, mobile phones, set-top boxes, speakers, etc. The electronic devices of this disclosure include all electronic devices that require a user to perform a desired operation by inputting specific commands.

[0045] Figure 1 It is a block diagram of an electronic device according to this disclosure.

[0046] Figure 1 The electronic device includes a memory 100, a processor 200, an output interface 300, and a microphone 400. The electronic device can be a device with both a display and memory, such as a TV or smartphone. Alternatively, the electronic device can be, for example, a device like an AI speaker, which has been commercially used in recent years, that recognizes a user's voice and provides desired information as sound.

[0047] The memory 100 includes a category database 120 and a content database 140.

[0048] Category database 120 is a pre-built database used to identify the content of user utterances and search for their meaning. Category database 120 includes the words to be identified, the categories set for each word, the instructions corresponding to each category, and a field for a score for each category. Figure 2 This is a diagram illustrating an example of category database 120. Category database 120 includes four fields: category field, word field, instruction field, and score field.

[0049] The word field includes every word that is expected to be included in the user's speech, and the category field includes a category name assigned to each of these words by classifying them by category. Figure 2 This displays the status for the categories <Music> and <Artist>. An instruction field is created for each category, containing the instructions to be executed for that category. Figure 2 This illustrates the situation where instructions for the <Music> category include "play," "open," and "reproduce." For example, the instruction "play" corresponds to the instruction to play the corresponding music. Therefore, for example, the word "AAA" and the instruction "play" are included in the utterance "play AAA," and the instruction "play" included in the <Music> category, which includes the word "AAA," exists in the category database 120. Thus, the speech recognition system can determine the user's utterance as a request to perform the operation of playing AAA.

[0050] Instructions may vary depending on the category. For example, for the <Music> category, instructions may primarily include playing music, and additionally, for example, instructions may include providing information related to the corresponding music. As another example, for the <Weather> category, instructions may primarily include requests for weather forecasts for a specific time or location, i.e., instructions such as "Let me know," "Tell me," and "How is it?" In this way, instructions are pre-set to match the characteristics of content included in the corresponding category on a category-by-category basis. Such instructions can be set by the manufacturer of the electronic device according to this disclosure. Furthermore, the user of the electronic device according to this disclosure can register his / her desired instructions in the category database 120.

[0051] The score field corresponds to each category and records the score for that category. The score for each category is incremented according to the processing described below in this disclosure. Figure 2 This shows the situation where 13 points and 5 points were assigned to the <Music> category and the <Singer> category, respectively.

[0052] Content database 140 stores various types of content that will be executed according to user commands. For example, content database 140 stores various sound sources (e.g., as included in...) Figure 2 The audio sources of the content in the word field (such as "AAA", "BBB", and "CCC") are stored as content corresponding to the <Music> category. Furthermore, in the presence of a <Movie> category, the video data for various movies is stored as content corresponding to that category.

[0053] On the other hand, in cases where information to be executed will be provided externally, the external server 500 retains the information. For example, in the case of weather information, a server operated by a meteorological agency or other weather information provider updates and provides the relevant information in real time. If the user wants a weather forecast, the electronic device receives the corresponding information from the external server 500.

[0054] The processor 200 is configured to execute a method for controlling speech recognition according to the present disclosure, and includes an execution unit 220, a learning unit 240, and a recognition unit 260.

[0055] A microphone 400 installed in an electronic device receives the user's voice and sends the received voice as an electrical signal to a recognition unit 260. The recognition unit 260, acting as a recognizer, queries a category database 120 for the received voice. Based on the query results, the recognition unit 250 classifies the voices recognized and those not recognized in the category database 120 into successful voices and unrecognized voices, respectively. For the unrecognized voices, the recognition unit 260 finds the instruction most likely corresponding to the user's intent and sends the found instruction to an execution unit 220 to perform an operation according to the corresponding instruction.

[0056] Learning unit 240 increases the score of the corresponding category in category database 120 for utterances classified as successful utterances by recognition unit 260. This increased score is used as a criterion for incorporating new instructions into instructions in future failed utterances.

[0057] The execution unit 220, acting as an actuator, performs operations corresponding to the user's speech based on the recognition result of the recognition unit 260. For example, the execution unit 220 reads the corresponding content stored in the content database 140 in the memory 100 to reproduce the specific content, and sends the read content to the output interface 300 for playback.

[0058] The output interface 300 includes a monitor 320 for displaying the operations performed by the execution unit 220 and a speaker 340. For example, when playing a movie, the monitor 320 is controlled by the execution unit 220 to display images, and the speaker 340 is controlled to output sound.

[0059] In the following text, reference will be made to Figure 3 and Figure 4 The operation of the electronic device having the above configuration according to this disclosure is described in detail.

[0060] Figure 3 and Figure 4 It is by Figure 1 A flowchart of a method for controlling speech recognition executed by an electronic device.

[0061] When the user speaks (S10), the spoken sound is received by the microphone 400 (S20).

[0062] The recognition unit 260 receives the signal of the user's speech from the microphone 400 and identifies the words and instructions included in the speech (S30). For example, when the user says "play yesterday", the recognition unit 260 identifies the words "yesterday" and the instruction "play".

[0063] The identification unit 260 queries the category database 120 to check if the identified instruction exists (S40). As a result of the check, when the instruction exists, the user's utterance is classified as a successful utterance, and when the instruction does not exist, the user's utterance is classified as a failed utterance (S50). For example, due to... Figure 2 The database contains the instruction "play", so the recognition unit 260 classifies the utterance "play yesterday" as a successful utterance. When the user's utterance is "I want to listen to yesterday", in... Figure 2 The instruction field does not contain the instruction "I want to hear", therefore the corresponding utterance is classified as a failure utterance.

[0064] At this point, the identification unit 260 can simultaneously consider the instructions and words included in the utterance to classify successful and unsuccessful utterances. That is, the identification unit 260 first queries the category database 120 to check the identified word categories and checks whether the instructions included in the utterance exist in the instruction field corresponding to the identified category. For example, the word "yesterday" included in the utterance "play yesterday" belongs to the <music> category. Therefore, the identification unit 260 determines whether the instruction "play" exists among the stored instructions corresponding to the <music> category.

[0065] Information about discourses classified as successful discourses is sent to learning unit 240. Learning unit 240 increments the category score for the words in category database 120, where the words are included in the discourses classified as successful discourses by recognition unit 260 (S100). For example, when the successful discourse "play yesterday" is created, learning unit 240 increments the current score of the <music> category for "yesterday" by 1. Figure 2 In the case of the <Music> category shown, the value of the score field increases from 13, which is the current score, to 14. Additionally, the recognition unit 260 sends a success message to the execution unit 220, and the execution unit 220 reads the corresponding data from the content database 140 to perform the operation corresponding to the success message and drives the output interface 300 to execute the corresponding instructions on the read data (S150). Therefore, the music "Yesterday" is played from the speaker 340.

[0066] For a failed utterance, the identification unit 260 determines the category of the words in the utterance (S110). For example, the utterance “I want to listen to yesterday” is classified as a failed utterance, and for a utterance classified as a failed utterance, the category of <music>, which includes “yesterday”, is determined as the category of the words in the utterance.

[0067] There are situations where multiple words exist within a failed utterance. For example, in the utterance "I want to listen to The Beatles' Yesterday," the word "The Beatles" is included in the <Artist> category, and the word "Yesterday" is included in the <Music> category. When multiple categories containing multiple words exist as described above, the identification unit 260 examines the category for each word and determines any one of the multiple categories as the category for the word in the utterance. At this time, the identification unit 260 determines the category with the highest score among the various categories as the category for the utterance.

[0068] The identification unit 260 selects any one of the instructions corresponding to the determined category as the instruction for the failed utterance (S120). The identification unit 260 checks which instruction is most frequently used in the corresponding category to select the instruction. For example, in the case of the <Music> category, if the instruction "play" is used most frequently, the identification unit 260 selects the instruction "play" as the instruction corresponding to the instruction "I want to listen" included in the failed utterance.

[0069] The recognition unit 260 requests the user to check the selected instruction (S130). The user's check request can be made by driving the output interface 300 through the execution unit 220. For example, the speaker 340 is driven to output the sound "Do you want to play yesterday?". As a result, in response to the user's failed statement "I want to hear yesterday", the speaker 340 outputs a request for the user to check whether he or she wants to play yesterday's sound. When the user responds to the output sound with an affirmative statement such as "yes" or "that's right", the recognition unit 260 receives the voice through the microphone 400 to check whether the recognition unit 260's determination matches the user's request (S140).

[0070] When the identification unit 260 confirms the result, it sends the result to the execution unit 220, and the execution unit 220 executes the result to play the music "Yesterday" through the speaker 340 (S150).

[0071] On the other hand, when the identification unit 260 determines the category for the failed speech in step S110, the learning unit 240 queries the score of the determined category (S210).

[0072] For example, regarding the instruction "I want to hear about yesterday" as a failure statement, in Figure 2 The score for the <Music> category, as determined by the identification unit 260, is 13. Therefore, the instruction "I want to listen" as a command in the failed discourse is added to the category database 120 as a new instruction for the <Music> category. At this time, in step S120, the identification unit 260 selects the newly added instruction and registers it as an instruction with the same operation as the instruction checked by the user in step S140. For example, the learning unit 240 adds the instruction "I want to listen" and gives the instruction the meaning of a playback operation instruction. As a result, the discourse including the words "I want to listen" in the <Music> category, which was previously classified as a failed discourse, is subsequently classified as a successful discourse.

[0073] On the other hand, when multiple words exist in a failed utterance, the instruction is registered for the category with the highest score among the categories of said multiple words. For example, when the failed utterance is "I want to listen to Yesterday of the Beatles," the utterance includes the words "The Beatles" as a <singer> category and the words "Yesterday" as a <music> category. Therefore, the identification unit 260 checks the <singer> category and the <music> category for the failed utterance. The learning unit 240 checks the scores of the categories. Figure 2 In the above, the score for the <Singer> category is 5, and the score for the <Music> category is 13. Therefore, learning unit 240 selects the <Music> category as the target to be registered for the instruction "I want to listen".

[0074] Based on the above processing, the learning unit 240 increases the score of the corresponding category for utterances classified as successful utterances. When this processing is repeated, the score of the corresponding category gradually increases to equal or greater than a preset value, and for such categories, instructions in failed utterances are newly added as valid instructions. Therefore, even if all instructions generated by various users were not previously recorded in the category database 120, new instructions are added to categories where many successful utterances have already appeared through machine learning methods.

[0075] A high score in a category indicates that the user frequently utters words and instructions within that category. Instructions set for any category within a category typically have the same or similar meanings. In the case of the <Music> category, for example, various instructions such as "play" and "open" mean "play" music. Therefore, if any category has been frequently targeted as a successful utterance, it means that words within that category are frequently used as targets for instructions with the same meaning. Thus, if words within a category are uttered after the user has frequently used the category, it can be assumed that even if the instruction in the utterance is not a pre-registered instruction, the instruction is one that is frequently executed for the corresponding category. Therefore, by processing speech recognition to appropriately execute intentions suitable for various users, the number of cases identified as failed utterances can be reduced.

[0076] Furthermore, even if the instruction included in the currently input utterance does not exist in the category database 120, the recognition unit 260 selects an instruction that matches the user's intent by querying the corresponding category. Therefore, even if an instruction is classified as a failed utterance, the instruction assumed to be the user's intent is selected and executed, without immediately resulting in recognition failure, and the user further examines the instruction. Thus, utterances identified as failed utterances are likely to lead to the same result as successful utterances.

[0077] Furthermore, words included in failed utterances can be words that do not exist in the category database 120. For example, in the case of the utterance "I want to hear hope," the word "hope" could be a word not included in any category. In this way, when words included in failed utterances do not exist in the category database 120, the learning unit 240 can create a new category in the category database 120 for words included in failed utterances. After creation, the same processing is applied to the new category created as described above.

[0078] The scores stored in the category database 120 of this disclosure represent a weight. The above embodiment illustrates a method of increasing the score by 1 point for each successful utterance, but it can be modified to a method where the increase in score for successful utterances gradually decreases as the number of successful utterances increases. Additionally, a method is provided to register instructions in failed utterances as new instructions in the category database 120 when the score of a successful utterance exceeds a preset value, but the preset value may not be set to a single value, but rather to a value across several stages. In this case, for example, for categories where the score exceeds the value of a first stage, the instructions for failed utterances are not immediately registered; the corresponding instructions are simply recorded in the category database 120, and the above embodiment can be modified to record the instructions in the category database 120 when the corresponding instructions are entered a specific number of times until the score exceeds the value of a second stage.

[0079] On the other hand, in this disclosure, the recognition unit 260 may be composed of modules that perform the following operations: performing speech recognition by converting the user's speech via speech-to-text (STT), checking categories through the category database 120, etc. Alternatively, the recognition unit 260 may be composed of separate modules for speech recognition and for querying the category database 120. Furthermore, in this case, the two hardware modules may be configured to be remotely spaced apart from each other, enabling data communication between them via wired or wireless communication networks. Additionally, the user's device may only have a microphone or only have a receiver 280 connected to the microphone, and some or all of the learning unit 240, recognition unit 260, and execution unit 220 may be configured to be located in a device or server capable of data communication with the user's device via a wired or wireless communication network.

[0080] Figures 5 to 7 These are illustrations showing other embodiments of the present disclosure, and illustrate various implementation examples of an electronic device equipped with the speech recognition system of the present disclosure. In the following embodiments, the same reference numerals are assigned to... Figure 1 The components shown in the embodiments are the same as those in the examples, and repeated descriptions are omitted. However, due to the... Figure 1 The same reference numerals for parts do not necessarily mean that the parts are exactly the same.

[0081] Figure 5 An implementation example of a method for the electronic device of this disclosure to communicate with a separate external device 410 is shown. Figure 1 In one embodiment, microphone 400 is included in the electronic device, and microphone 400 digitizes analog voice signals and sends the digitized analog voice signals to the processor of the electronic device. However, in Figure 5In this embodiment, microphone 400a is housed in a separate external device 410, rather than within a component included in the electronic device. Microphone 400a digitizes analog voice signals, and external device 410 transmits the digitized voice signals to the electronic device via communication methods such as Bluetooth, infrared communication, and WiFi. Figure 5 In the embodiments, such as Figure 1 The microphone 400 shown can be installed inside the electronic device together with the microphone 400a in the external device 410.

[0082] exist Figure 5 In the example, external device 410 can be, for example, a device such as a remote control, and the corresponding electronic device can be, for example, a TV. In this case, the remote control includes, for example, using a smartphone as a remote control by installing a remote control app on the smartphone capable of controlling the TV. As another example, external device 410 can be a device that performs functions equivalent to those of the electronic device, rather than a device attached to the electronic device, such as a remote control. For example, both the electronic device and the external device can be a smartphone; the electronic device can be a TV, and the external device can be a smartphone.

[0083] The electronic device also includes a receiver 280 capable of communicating with an external device 410. Communication between the external device 410 and the receiver 280 can be performed in various ways (e.g., Bluetooth, infrared communication, communication via a Wi-Fi network, wired communication, etc.). When a user speaks to the external device 410, the microphone 400a of the external device 410 receives the speech and sends it to the receiver 280, which then sends the speech to the recognition unit 260. The voice recognition control processing is the same as in the embodiment described above.

[0084] On the other hand, in this embodiment, the external device 410 may be a device such as a smartphone or AI speaker used by the user, and the electronic device itself may be an externally configured server. For example, when a user uses an AI speaker to control interactive network television (IPTV) in their home by voice, the AI ​​speaker is equipped with a microphone 400a as an external device 410 to receive the user's speech, and the electronic device may consist of the AI ​​speaker and a server of an operator connected via a wired or wireless network to perform voice recognition at a remote location.

[0085] According to this embodiment, the microphone 400a that directly receives the user's speech and the electronic device that receives the sound signal transmitted from the microphone 400a to perform the processing of this disclosure are manufactured as separate, independent devices. Therefore, the operation of the electronic device can be controlled by the user's speech at a remote location.

[0086] Figure 6An implementation example of a method for the electronic device of this disclosure to communicate with a separate output device 300a without excluding the output interface 300 is shown. Figure 1 In some embodiments, the output interface 300 is included in the electronic device. However, in Figure 6 In some embodiments, the output device 300a is housed in a separate external device, rather than in a component included within the electronic device. For example, in Figure 1 The output interface 300 shown can also be set in Figure 6 In electronic devices.

[0087] The output device 300a includes a monitor 320a and a speaker 340a. Here, the electronic device can be, for example, a device such as a smartphone, and the corresponding output device 300a can be a device such as a TV. As another example, the electronic device can be composed of an AI speaker, and the output device can be composed of a TV. In this case, the speaker 340 can be disposed inside the electronic device, and the monitor 320a can be disposed inside the output device 300a. Alternatively, the monitor 320a can be configured to be disposed inside the electronic device, and the speaker 340a can be configured to be disposed outside the electronic device.

[0088] The electronic device includes an interface 210. Interface 210 provides a communication port for connection to the output device 300a. Interface 210 can be a port capable of simultaneously transmitting video / audio signals, such as HDMI, DP, and Thunderbolt. Furthermore, interface 210 can be composed of ports through which video and audio signals are transmitted to different paths.

[0089] Figure 7 An implementation example of a method for the electronic device of this disclosure to communicate with separate servers 120a and 140a without excluding memory 100 is shown. Figure 1 In some embodiments, the memory 100 is included in the electronic device. However, in Figure 7 In some embodiments, the memory is located in a separate external device, rather than in a component included in the electronic device. The memory may also be located in... Figure 7 Inside electronic devices.

[0090] exist Figure 7In this embodiment, the memory includes a category database server 120a for storing a category database and a content database server 140a for storing a content database. The electronic device includes a communication interface 290 that serves as a communication function between servers 120a and 140a and processor 200. The communication interface 290 can communicate with servers 120a and 140a using communication methods such as wired / wireless internet networks or Bluetooth.

[0091] When the electronic device of this disclosure is configured as, for example, an AI speaker or a set-top box, the data of the content to be executed by the execution unit 220 can be stored in an external content database server 140a connected via a wired or wireless network. Typically, operators providing IPTV services store a database of the content to be provided in a separate content database server 140a connected via a wired communication network, and when a user selects, for example, a specific movie by manipulating the IPTV, the data of that movie is transmitted to the user's IPTV. In such a system, for example, when the set-top box or AI speaker attached to the IPTV is configured as the electronic device of this disclosure, the content to be reproduced can be received from the operator's content database server 140a for playback on the IPTV display.

[0092] As in Figure 1 As shown, the electronic device may store the category database 120 in a memory 100 provided in the electronic device, but it may also store the category database 120 in a memory such as a memory 100 provided in the electronic device. Figure 7 The external category database server 120a is shown. Category database 120 is typically manufactured by the electronic device manufacturer, and it is continuously updated as users use the electronic device of this disclosure, with new instructions added as identifiable instructions. When the updated category database 120 is configured differently for each user, it is preferably stored in the memory 100 located within the electronic device. Thus, each user can have a dedicated category database 120 for themselves. Even if the category database 120 is configured differently for each user, it can still be stored in the external category database server 120a. When the category database 120 is configured to be updated based on usage results from several users, it is preferably stored in the external category database server 120a. Even in this case, a method can be employed where the category database 120 is located in the memory 100 of the electronic device and an external category database server 120a is also located.

[0093] Figure 8 The electronic device shown in this disclosure includes other external devices D1, D2, ..., D nExample of a communication interface 295 for communication. Communication interface 295 provides a way for other external devices D1, D2, ..., D... n It enables communication with electronic devices via means such as Bluetooth, Wi-Fi, and infrared communication. Here, external devices D1, D2, ..., D... n This could be a display device, such as a TV. As an example, when the electronic device of this disclosure is a smartphone, the results of this disclosure implemented by the smartphone can be shared with a TV via Bluetooth communication. The communication interface 295 can be configured to perform communication with external devices D1, D2, ..., D... n It can be used for 1:1 communication, or it can be configured to implement 1:n communication.

[0094] Although exemplary embodiments of this disclosure have been shown and described above, this disclosure is not limited to the specific exemplary embodiments described above, but various modifications can be made by those skilled in the art to which this disclosure pertains without departing from the scope and spirit of this disclosure as disclosed in the appended claims. Such modifications should also be understood to fall within the scope of this disclosure.

Claims

1. An electronic device, comprising: The processor is configured to execute one or more programs stored in memory to perform the following operations: Based on the presence of instructions included in the user's voice input in the database, scores are added for categories corresponding to the words included in the user's voice input; as well as Since the instruction included in the user's voice input does not exist in the database, the score of the category corresponding to the word included in the user's voice input is checked to see if it is equal to or greater than a preset value. Based on the check indicating that the score is equal to or greater than the preset value, the instruction is registered in the database, such that the instruction is included in the category corresponding to the word. The score for the category increases cumulatively based on the user's voice input.

2. The electronic device according to claim 1, in, Based on the fact that the instruction does not exist in the database and the user's voice input includes multiple words, the processor is also configured to execute the one or more programs stored in the memory to perform the following operations: Examine the category corresponding to each of the multiple words in the multiple categories; as well as The category among the multiple categories is identified based on the scores of the multiple categories being examined.

3. The electronic device according to claim 1, in, The processor is also configured to execute one or more programs stored in the memory to perform the following operations: creating a new category for the words included in the user's voice input based on the fact that the words included in the user's voice input do not exist in the database.

4. The electronic device according to claim 1, in, Based on the fact that the instruction does not exist in the database and that the category corresponding to the word includes multiple instructions, the processor is also configured to execute the one or more programs stored in the memory to perform the following operation: selecting one of the multiple instructions as the instruction for the word.

5. The electronic device according to claim 4, in, Since the instruction does not exist in the database, and the user's voice input includes multiple words and each of the multiple categories corresponds to a word among the multiple words, the processor is also configured to execute the one or more programs stored in the memory to perform the following operation: selecting a category among the multiple categories based on the scores of the multiple categories.

6. The electronic device according to claim 4, in, The processor is also configured to execute one or more programs stored in the memory to perform the following operations: receiving user checks for selected instructions.

7. The electronic device according to claim 1, further comprising: A memory is configured to store the database, and the database includes various fields for the words, the instructions, the categories, and the scores.

8. The electronic device according to claim 1, further comprising: The communication interface is configured to perform communication for querying the database on an external database server where the database is stored.

9. The electronic device according to claim 1, further comprising: An executor is configured to perform operations corresponding to the words and instructions.

10. The electronic device according to claim 9, further comprising: The output interface is configured to output the result of the operation performed by the actuator.

11. The electronic device according to claim 9, further comprising: An interface is configured to send the result of the operation performed by the actuator to an external output device, such that the external output device displays the result of the operation.

12. The electronic device according to claim 1, further comprising: The microphone is configured to receive the sound of the user's voice input and send the received sound to the recognizer.

13. The electronic device according to claim 1, further comprising: The receiver is configured to receive a signal for the sound from an external device including a microphone that receives the sound of the user's voice input, and to send the received signal to a recognizer.

14. A method for voice recognition for controlling electronic devices, comprising: Based on the presence of instructions included in the user's voice input in the database, scores are added for categories corresponding to the words included in the user's voice input; and Since the instruction included in the user's voice input does not exist in the database, the score of the category corresponding to the word included in the user's voice input is checked to see if it is equal to or greater than a preset value. Based on the check indicating that the score is equal to or greater than the preset value, the instruction is registered in the database, such that the instruction is included in the category corresponding to the word. The score for the category increases cumulatively based on the user's voice input.

15. The method according to claim 14, in, The method further includes the following based on the fact that the instruction does not exist in the database and the user's voice input includes multiple words: The category corresponding to each of the multiple words is identified based on scores in multiple categories.

Images