Electronic device and control method thereof

Abstract

An electronic device and a control method thereof are provided. The control method of the electronic device includes: receiving a first artificial intelligence model and a second artificial intelligence model used by the first external electronic device and the second external electronic device, and a plurality of learning data stored in the first external electronic device and the second external electronic device from the first external electronic device and the second external electronic device, respectively; identifying first learning data corresponding to the second learning data received from the second external electronic device among the learning data received from the first external electronic device; training the second artificial intelligence model used by the second external electronic device based on the first learning data; and sending the trained second artificial intelligence model to the second external electronic device.

Description

Technical Field

[0001] The present disclosure relates to an electronic device and a control method. More specifically, the present disclosure relates to a method for training an artificial intelligence model of at least one external electronic device among a plurality of external electronic devices based on learning data stored in the plurality of external electronic devices.

[0002] In addition, the present disclosure relates to an artificial intelligence (AI) system and its application that utilizes a machine learning algorithm to simulate functions such as cognition and determination of the human brain. Background Art

[0003] Artificial intelligence (AI) systems are computer systems that achieve human-level intelligence and are systems in which machines learn, determine, and become more intelligent on their own, unlike rule-based intelligent systems. Artificial intelligence systems show higher recognition rates as they are used more and become able to more correctly understand user preferences. For this reason, rule-based intelligent systems are gradually being replaced by deep learning-based artificial intelligence systems.

[0004] AI technologies include machine learning (deep learning) and meta-technologies that leverage machine learning.

[0005] Machine learning refers to algorithmic technology that classifies / learns the characteristics of the input data itself. Meanwhile, meta-technology refers to technology that utilizes machine learning algorithms such as deep learning, and includes technical fields such as language understanding, visual understanding, reasoning / prediction, knowledge representation, and operation control.

[0006] Examples of various fields where artificial intelligence technology is applied are as follows. Language understanding refers to the technology of recognizing human language / characters and applying / processing them, and includes natural speech processing, machine translation, communication systems, query and response, speech recognition / synthesis, etc. Visual understanding refers to the technology of recognizing objects and processing objects in a manner similar to human vision, and includes object recognition, object tracking, image search, human recognition, scene understanding, spatial understanding, image improvement, etc. Reasoning / prediction refers to the technology of determining information and then performing logical reasoning and prediction, including knowledge / probability-based reasoning, optimization prediction, preference-based planning, recommendation, etc. Knowledge representation refers to the technology of automatically processing information experienced by humans into knowledge data, including knowledge construction (data generation / classification), knowledge management (data utilization), etc. Operation control refers to the technology of controlling vehicle autonomous driving and robot movement, including motion control (navigation, collision, driving), operation control (behavior control), etc.

[0007] Meanwhile, various methods of using AI models in devices with small resources such as user devices are being discussed recently. In addition, a method of building a personalized AI model by training the AI ​​model with data stored in the user device as learning data is also being discussed.

[0008] Recently, users have various types of user devices. Various user devices, such as smart phones, artificial intelligence speakers, digital TVs, refrigerators, etc., may include AI models. However, there is a problem that the amount of learning data collected by some user devices is not enough and is not enough to train the AI ​​model, or the learning level is different for each user device, so even if the same user uses the user device, the same final performance value cannot be maintained.

[0009] The above information is presented as background information only to assist with an understanding of the present disclosure. No determination has been made, and no assertion is made, as to whether any of the above might be applicable as prior art with respect to the present disclosure. Summary of the invention

[0010] Workaround

[0011] An aspect of the present disclosure is to solve at least the above problems and / or disadvantages and to provide at least the advantages described below. Therefore, one aspect of the present disclosure is to solve the above problems and is directed to a method of sharing learning data between a plurality of external electronic devices.

[0012] Additional aspects will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the presented embodiments.

[0013] According to one aspect of the present disclosure, a control method for an electronic device for achieving the above-mentioned purpose is provided. The control method includes: receiving a first artificial intelligence model and a second artificial intelligence model used by a first external electronic device and a second external electronic device, and a plurality of learning data stored in the first external electronic device and the second external electronic device from a first external electronic device and a second external electronic device, respectively; identifying first learning data corresponding to the second learning data received from the second external electronic device among the learning data received from the first external electronic device; training the second artificial intelligence model used by the second external electronic device based on the first learning data; and sending the trained second artificial intelligence model to the second external electronic device.

[0014] The control method may also include: respectively receiving first characteristic information of the first external electronic device and second characteristic information of the second external electronic device; converting the first learning data into third learning data for training a second artificial intelligence model used by the second external electronic device based on the first characteristic information of the first external electronic device and the second characteristic information of the second external electronic device; and training the second artificial intelligence model used by the second external electronic device based on the third learning data.

[0015] Furthermore, at the time of identification, at least one of the input value and the label value included in the second learning data may be compared with the input value and the label value included in the learning data received from the first external electronic device, and the first learning data may be acquired.

[0016] The second artificial intelligence model may be an artificial intelligence model for speech recognition, and the plurality of learning data may include speech data, label values ​​of the speech data, and user information corresponding to the speech data.

[0017] In addition, when performing recognition, at least one of the second voice data, the second label value of the second voice data, and the second user information corresponding to the second voice data included in the second learning data can be compared with the first voice data, the first label value of the first voice data, and the first user information corresponding to the first voice data included in the learning data received from the first external electronic device, and the first learning data can be obtained.

[0018] The first characteristic information and the second characteristic information may include at least one of the following items: characteristic information related to a voice input device included in the first external electronic device and the second external electronic device, characteristic information related to noise input through the voice input device of the first external electronic device and the second external electronic device, and characteristic information related to a distance between a location where a user voice is generated and a location of the first external electronic device and the second external electronic device.

[0019] Meanwhile, at the time of conversion, the speech data included in the first learning data may be converted by using a frequency filter.

[0020] In addition, when receiving, based on predetermined time conditions, the first artificial intelligence model and the second artificial intelligence model used by the first external electronic device and the second external electronic device, a plurality of learning data stored in the first external electronic device and the second external electronic device, and the first feature information of the first external electronic device and the second feature information of the second external electronic device can be received respectively.

[0021] At the same time, the electronic device according to an embodiment of the present disclosure includes a memory, a communication device and a processor, and the processor is configured to receive a first artificial intelligence model used by the first external electronic device and a second artificial intelligence model used by the second external electronic device, as well as a plurality of learning data stored in the first external electronic device and the second external electronic device from the first external electronic device and the second external electronic device respectively via the communication device; identify first learning data corresponding to the second learning data received from the second external electronic device among the learning data received from the first external electronic device; train the second artificial intelligence model used by the second external electronic device based on the first learning data; and send the trained second artificial intelligence model to the second external electronic device via the communication device.

[0022] Here, the processor can receive first characteristic information of the first external electronic device and second characteristic information of the second external electronic device respectively via the communication device, convert the first learning data into third learning data for training the second artificial intelligence model used by the second external electronic device based on the first characteristic information of the first external electronic device and the second characteristic information of the second external electronic device, and train the second artificial intelligence model used by the second external electronic device based on the third learning data.

[0023] Furthermore, the processor may compare at least one of the input value and the label value included in the second learning data with the input value and the label value included in the learning data received from the first external electronic device, and acquire the first learning data.

[0024] The second artificial intelligence model may be an artificial intelligence model for speech recognition, and the plurality of learning data may include speech data, label values ​​of the speech data, and user information corresponding to the speech data.

[0025] In addition, the processor can compare at least one of the second voice data, the second label value of the second voice data, and the second user information corresponding to the second voice data included in the second learning data with the first voice data, the first label value of the first voice data, and the first user information corresponding to the first voice data included in the learning data received from the first external electronic device, and obtain the first learning data.

[0026] The first characteristic information and the second characteristic information may include at least one of the following items: characteristic information related to a voice input device included in the first external electronic device and the second external electronic device, characteristic information related to noise input through the voice input device of the first external electronic device and the second external electronic device, and characteristic information related to a distance between a location where a user voice is generated and a location of the first external electronic device and the second external electronic device.

[0027] Furthermore, the processor may convert the voice data included in the first learning data by using a frequency filter.

[0028] Based on predetermined time conditions, the processor can receive, through a communication device, a first artificial intelligence model used by a first external electronic device and a second artificial intelligence model used by a second external electronic device, a plurality of learning data stored in the first external electronic device and the second external electronic device, and first feature information of the first external electronic device and second feature information of the second external electronic device.

[0029] At the same time, according to an embodiment of the present disclosure, an electronic device includes a memory, a communication device and a processor, and the processor is configured to receive multiple learning data stored in an external electronic device from the external electronic device via the communication device, identify first learning data corresponding to second learning data stored in the electronic device among the multiple learning data received from the external electronic device, and train an artificial intelligence model used by the electronic device based on the identified first learning data.

[0030] The processor can receive characteristic information of an external electronic device via a communication device, convert the first learning data into third learning data for training an artificial intelligence model used by the electronic device based on the characteristic information of the external electronic device and the characteristic information of the electronic device, and train the artificial intelligence model used by the electronic device based on the third learning data.

[0031] Furthermore, the processor may compare at least one of an input value and a label value included in the second learning data with input values ​​and label values ​​included in a plurality of learning data received from the external electronic device, and identify the first learning data.

[0032] The characteristic information of the electronic device and the external electronic device may include at least one of the following items: characteristic information related to the voice input device of the electronic device and the external electronic device, characteristic information related to the noise input through the voice input device of the electronic device and the external electronic device, and characteristic information related to the distance between the location where the user voice is generated and the location of the electronic device and the external electronic device.

[0033] According to the above-described various embodiments of the present disclosure, the electronic device may provide learning data suitable for an external electronic device for which personalized learning data is insufficient, and train an artificial intelligence model by using the provided learning data.

[0034] Other aspects, advantages, and salient features of the disclosure will become apparent to those skilled in the art from the following detailed description, which, taken in conjunction with the annexed drawings, discloses various embodiments of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The above and other aspects, features and advantages of certain embodiments of the present disclosure will become more apparent from the following description in conjunction with the accompanying drawings, in which:

[0036] Figure 1 is a diagram schematically illustrating an embodiment of the present disclosure;

[0037] Figure 2 is a block diagram showing a schematic configuration of an electronic device according to an embodiment of the present disclosure;

[0038] Figure 3 is a block diagram showing a detailed configuration of an electronic device according to an embodiment of the present disclosure;

[0039] Figure 4 is a diagram illustrating a data sharing method according to an embodiment of the present disclosure;

[0040] Figure 5 is a diagram illustrating a data sharing method according to an embodiment of the present disclosure;

[0041] Figure 6 is a diagram illustrating a method of converting learning data according to an embodiment of the present disclosure;

[0042] Figure 7 is a diagram illustrating a method of acquiring learning data according to an embodiment of the present disclosure;

[0043] Figure 8 is a diagram illustrating a method of sharing learning data according to an embodiment of the present disclosure;

[0044] Fig. 9 is a diagram illustrating a method for a second external electronic device to convert learning data according to an embodiment of the present disclosure; and

[0045] Fig.10 is a flowchart illustrating a control method of an electronic device according to an embodiment of the present disclosure.

[0046] Throughout the drawings, like reference numerals will be understood to refer to like parts, components and structures. DETAILED DESCRIPTION

[0047] The following description with reference to the accompanying drawings is provided to assist in a comprehensive understanding of the various embodiments of the present invention as defined by the claims and their equivalents. Various specific details are included to assist in understanding, but these are considered to be exemplary only. Therefore, those of ordinary skill in the art will recognize that various changes and modifications may be made to the various embodiments described herein without departing from the scope and spirit of the present disclosure. In addition, descriptions of well-known functions and structures may be omitted for clarity and brevity.

[0048] The terms and words used in the following description and claims are not limited to the dictionary meanings, but are used only by the inventor to enable a clear and consistent understanding of the present disclosure. Therefore, it should be clear to those skilled in the art that the following description of various embodiments of the present disclosure is provided for illustrative purposes only and not for limiting the disclosure defined by the attached claims and their equivalents.

[0049] It is to be understood that the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a component surface" includes reference to one or more of such surfaces.

[0050] In the present disclosure, expressions such as “having”, “may have”, “including”, and “may include” should be interpreted as indicating the presence of such features (e.g., elements such as values, functions, operations, and components), and the terminology is not intended to exclude the presence of additional features.

[0051] In addition, in the present disclosure, the expression "A or B", "at least one of A and / or B", or "one or more of A and / or B", etc. may include all possible combinations of the listed items. For example, "A or B", "at least one of A and B", or "at least one of A or B" refers to all of the following situations: (1) including at least one A, (2) including at least one B, or (3) including at least one A and at least one B.

[0052] In addition, the expressions "first", "second", etc. used in the present disclosure may be used to describe various elements, regardless of any order and / or importance. In addition, such expressions are only used to distinguish one element from another element, rather than to limit these elements.

[0053] Furthermore, a description in the present disclosure that one element (e.g., a first element) is “(operably or communicatively) coupled” or “connected to” another element (e.g., a second element) should be interpreted to mean that the one element may be directly coupled to the other element, or the one element may be coupled to the other element through another element (e.g., a third element). Conversely, a description that one element (e.g., a first element) is “directly coupled” or “directly connected” to another element (e.g., a second element) may be interpreted to mean that another additional element (e.g., a third element) does not exist between the one element and the other element.

[0054] In addition, the expression "configured to" used in the present disclosure may be used interchangeably with other expressions, such as "suitable for", "capable of", "designed to", "suitable for", "made of", and "capable of", depending on the circumstances. At the same time, the term "configured to" does not necessarily mean that the device is "specially designed to" in terms of hardware. On the contrary, in some cases, the expression "the device is configured to" may mean that the device is "capable of" performing an operation together with another device or component. For example, the phrase "a sub-processor configured to perform A, B, and C" may mean a dedicated processor (e.g., an embedded processor) for performing the corresponding operations, or a general-purpose processor (e.g., a central processing unit (CPU) or an application processor) that can perform the corresponding operations by executing one or more software programs stored in a memory device.

[0055] The electronic device according to various embodiments of the present disclosure may include, for example, a smart phone, a tablet personal computer (PC), a mobile phone, a video phone, an e-book reader, a desktop PC, a laptop PC, a netbook computer, a workstation, a server, a personal digital assistant (PDA), a portable multimedia player (PMP), a Moving Picture Experts Group Stage 1 or Stage 2 (MPEG-1 or MPEG-2) Audio Layer 3 (MP3) player, a medical device, a camera, or at least one of a wearable device. Meanwhile, the wearable device may include an accessory device (e.g., a watch, a ring, a bracelet, an anklet, a necklace, glasses, contact lenses, or a head mounted device (HMD)), a device integrated with fabric or clothing (e.g., electronic clothing), a body attachment device (e.g., a skin pad or a tattoo), or an implantable circuit. In addition, in some embodiments, the electronic device according to various embodiments of the present disclosure may include, for example, a television, a digital versatile disk (DVD) player, a speaker, a refrigerator, an air conditioner, a cleaner, an oven, a microwave oven, a washing machine, an air cleaner, a set-top box, a home automation control panel, a security control panel, a media box (e.g., a Samsung apple or Google ), game consoles (e.g. ), at least one of an electronic dictionary, an electronic key, a camera or an electronic photo frame.

[0056] According to another embodiment of the present disclosure, the electronic device may include various types of medical instruments (for example, various types of portable medical measuring instruments (blood glucose meters, heart rate meters, blood pressure meters or thermometers, etc.), magnetic resonance angiography (MRA), magnetic resonance imaging (MRI), computed tomography (CT), photographic equipment, ultrasound instruments, etc.), navigation equipment, global navigation satellite systems (GNSS), event data recorders (EDRs), flight data recorders (FDRs), vehicle infotainment equipment, electronic devices for ships (for example, navigation equipment for ships, gyroscopes, etc.), avionics devices, security equipment, head units for vehicles, industrial or household robots, drones, automatic teller machines (ATMs) of financial institutions, sales terminals (POS) of stores, or Internet of Things (IoT) devices (for example, light bulbs, various types of sensors, sprinkler equipment, fire alarms, thermostats, street lights, ovens, exercise equipment, hot water tanks, heaters, boilers, etc.).

[0057] Furthermore, in the present disclosure, the term “user” may refer to a person using an electronic device or a device using an electronic device (eg, an artificial intelligence electronic device).

[0058] Meanwhile, in the present disclosure, the first artificial intelligence model may mean an artificial intelligence model used by a first external electronic device or an artificial intelligence model received from the first external electronic device, and the second artificial intelligence model may mean an artificial intelligence model used by a second external electronic device or an artificial intelligence model received from the second external electronic device.

[0059] Meanwhile, in the present disclosure, the first learning data may mean learning data stored in the first external electronic device, and the second learning data may mean learning data stored in the second external electronic device.

[0060] Hereinafter, the present disclosure will be described in more detail with reference to the accompanying drawings.

[0061] Figure 1 It is a diagram schematically illustrating an embodiment of the present invention.

[0062] Reference Figure 1, multiple external electronic devices (e.g., a first external electronic device 200-1, a second external electronic device 200-2, and a third external electronic device 200-3) can communicate with the electronic device 100. The multiple external electronic devices may include an artificial intelligence model, and may be a device for providing various services by using an artificial intelligence model. Here, the artificial intelligence model used by the multiple external electronic devices may be different according to the purpose of each of the multiple external electronic devices. For example, the artificial intelligence model may be different, such as an artificial intelligence model for speech recognition, an artificial intelligence model for image analysis, and the like. The electronic device 100 is a device for training artificial intelligence models used by multiple external electronic devices.

[0063] Specifically, the electronic device 100 can store artificial intelligence models used by multiple external electronic devices and data necessary for training the artificial intelligence models. Specifically, the electronic device 100 can receive artificial intelligence models of each of the multiple external electronic devices from the multiple external electronic devices and store these models. Here, the artificial intelligence model of each of the multiple external electronic devices can be an artificial intelligence model trained based on learning data of each of the multiple external electronic devices. However, the present disclosure is not limited to this, and the electronic device 100 can receive the artificial intelligence model used by each of the multiple external electronic devices from an external server and store the model.

[0064] At the same time, the electronic device 100 can receive learning data from each of the multiple external electronic devices and store the data. Here, the electronic device 100 can classify the artificial intelligence model and learning data received from each of the multiple external electronic devices according to each external electronic device and store them.

[0065] The electronic device 100 can train the artificial intelligence model used by each of the multiple external electronic devices based on the learning data received from each of the multiple external electronic devices. For example, the electronic device 100 can train the first artificial intelligence model used by the first external electronic device 200-1 based on the learning data received from the first external electronic device 200-1 to the third external electronic device 200-3. By the same method, the electronic device 100 can train the second artificial intelligence model used by the second external electronic device 200-2 based on the learning data received from the first external electronic device 200-1 to the third external electronic device 200-3. Similarly, by the same method, the electronic device 100 can train the third artificial intelligence model used by the third external electronic device 200-3 based on the learning data received from the first external electronic device 200-1 to the third external electronic device 200-3.

[0066] Through the above method, multiple external electronic devices can share the learning data stored in each device and build a personalized artificial intelligence model. In addition, for example, the first external electronic device 200-1 is a device in which a large amount of learning data is stored. In addition, in the case where the learning data stored in the second external electronic device 200-2 and the third external electronic device 200-3 is smaller than that of the first external electronic device 200-1, the following effect will be obtained, that is, the second external electronic device 200-2 and the third external electronic device 200-3 can train the artificial intelligence model by using the learning data stored from the first external electronic device 200-1 to solve the problem of insufficient learning data.

[0067] At the same time, the artificial intelligence models that share learning data as described above may be artificial intelligence models that perform similar functions. For example, all of the first to third artificial intelligence models may be artificial intelligence models related to speech recognition. Alternatively, all of the first to third artificial intelligence models may be artificial intelligence models for image analysis. However, the present disclosure is not limited thereto, and in the case where learning data needs to be shared, learning data stored by different artificial intelligence models may be shared by multiple external electronic devices.

[0068] At the same time, the electronic device 100 can train the artificial intelligence models of multiple external electronic devices by using an external server storing various learning data. In this case, training the artificial intelligence model by using all the learning data of the external server has the disadvantage of requiring a large amount of resources. Therefore, the electronic device 100 can select learning data similar to the learning data included in the first external electronic device 200-1 to the third external electronic device 200-3 from a large amount of learning data stored by the external server, and train the artificial intelligence model based on the selected learning data.

[0069] Hereinafter, various embodiments of the present disclosure will be described based on the first external electronic device 200-1 and the second external electronic device 200-2 among the plurality of external electronic devices. Meanwhile, the technical idea of ​​the present disclosure may be applied to two or more external electronic devices.

[0070] Figure 2 is a block diagram showing a schematic configuration of an electronic device according to an embodiment of the present disclosure.

[0071] The electronic device 100 may include a memory 110 , a communication device 120 , and a processor 130 .

[0072] The memory 110 may store instructions or data related to at least one other different component of the electronic device 100. In particular, the memory 110 may be implemented as a non-volatile memory, a volatile memory, a flash memory, a hard disk drive (HDD) or a solid state drive (SSD), etc. The memory 110 may be accessed by the processor 130, and data reading / recording / correction / deletion / update, etc. may be performed by the processor 130. In the present disclosure, the term "memory" may include the memory 110, a read-only memory (ROM) (not shown) and a random access memory (RAM) (not shown) within the processor 130, or a memory card (not shown) (e.g., a micro secure digital (SD) card, a memory stick) installed on the electronic device 100.

[0073] In addition, the memory 110 may receive the artificial intelligence model used by the first external electronic device 200-1 and the second external electronic device 200-2, and the learning data stored in the first external electronic device 200-1 and the second external electronic device 200-2 from the first external electronic device 200-1 and the second external electronic device 200-2. In addition, the memory 110 may receive information about the feature information of the first external electronic device 200-1 and the second external electronic device 200-2, and store the information. Here, the feature information of the first external electronic device 200-1 and the second external electronic device 200-2 may include at least one of the following items: feature information related to the voice input device included in the first external electronic device 200-1 and the second external electronic device 200-2, feature information related to the noise input through the voice input device of the first external electronic device 200-1 and the second external electronic device 200-2, and feature information related to the distance between the position where the user voice is generated and the first external electronic device 200-1 and the second external electronic device 200-2.

[0074] The communication device 120 is a component for performing communication with another electronic device. At the same time, the communication connection of the communication device 120 with another electronic device may include communication through a third device (e.g., a repeater, a hub, an access point, a server, a gateway, etc.). Wireless communication may include cellular communication using at least one of Long Term Evolution (LTE), LTE Advanced (LTE-A), Code Division Multiple Access (CDMA), Wideband CDMA (WCDMA), Universal Mobile Telecommunications System (UMTS), Wireless Broadband (WiBro), or Global System for Mobile Communications (GSM). According to an embodiment, wireless communication may include, for example, Wi-Fi, Bluetooth, Bluetooth Low Energy (BLE), Zigbee, Near Field Communication (NFC), Magnetic Security Transmission, Radio Frequency (RF), or Body Area Network (BAN). Wired communication may include, for example, at least one of Universal Serial Bus (USB), High Definition Multimedia Interface (HDMI), Recommended Standard 232 (RS-232), Power Line Communication, or Plain Old Telephone Service (POTS). The network performing wireless communication or wired communication may include at least one of a telecommunication network such as a computer network (eg, a local area network (LAN) or a wide area network (WAN)), the Internet, or a telephone network.

[0075] Specifically, the communication device 120 may receive the artificial intelligence models used by the first external electronic device 200-1 and the second external electronic device 200-2, and the learning data stored in the first external electronic device 200-1 and the second external electronic device 200-2 from the first external electronic device 200-1 and the second external electronic device 200-2.

[0076] The processor 130 may be electrically connected to the memory 110 and control overall operations and functions of the electronic device 100 .

[0077] Specifically, the processor 130 may identify first learning data corresponding to second learning data received from the second external electronic device 200-2 among a plurality of learning data received from the first external electronic device 200-1, and train an artificial intelligence model used by the second external electronic device 200-2 based on the identified first learning data. In this case, the first external electronic device 200-1 may be an external server including a large amount of learning data, and the second external electronic device 200-2 may be a user terminal.

[0078] At the same time, the processor 130 may receive feature information of each of the first external electronic device 200-1 and the second external electronic device 200-2 from the first external electronic device 200-1 and the second external electronic device 200-2. The electronic device 100 may convert the first learning data into third learning data for training the second artificial intelligence model based on the feature information of the first external electronic device 200-1 and the second external electronic device 200-2.

[0079] For example, in the case where the second artificial intelligence model is an artificial intelligence model for speech recognition, the learning data may be data of a user voice. Here, the user voice received by the first external electronic device 200-1 and the second external electronic device 200-2 may be different according to the characteristic information of each external electronic device. For example, it may be assumed that the first external electronic device 200-1 is a user terminal such as a smart phone, and the second external electronic device 200-2 is a refrigerator. The first external electronic device 200-1 and the second external electronic device 200-2 may be different from each other in terms of the type of microphones included inside the device, the number of microphones, etc., and the distance from the starting point of the user voice to the external electronic device may also be different. Therefore, the noise, frequency characteristics, etc. included in the user voice received by the first external electronic device 200-1 may be different from the noise, frequency characteristics, etc. included in the user voice received by the second external electronic device 200-2. Therefore, in the case where the electronic device 100 is to use the second artificial intelligence model, it is necessary to convert the first learning data to adapt to the characteristics of the second external electronic device 200-2. Therefore, the processor 130 may convert the first learning data into third learning data for training the second artificial intelligence model based on the feature information of the first external electronic device 200 - 1 and the second external electronic device 200 - 2 .

[0080] In the above embodiment, the case where the artificial intelligence model is used for speech recognition is described. However, in the case of artificial intelligence models that perform different functions, the same technical ideas can be applied as needed.

[0081] At the same time, the processor 130 may compare at least one of the input value and the label value included in the second learning data with the input value and the label value included in each of the multiple learning data received from the first external electronic device 200-1, and obtain the first learning data. Specifically, in the case where the first external electronic device 200-1 is an external server including a large amount of learning data, the electronic device 100 can identify the learning data used to train the second artificial intelligence model among the learning data of the first external electronic device 200-1. In this case, the electronic device 100 can identify learning data similar to the input value and label value of the second learning data from among the multiple learning data of the first external electronic device 200-1.

[0082] For example, in the case where the first artificial intelligence model and the second artificial intelligence model are artificial intelligence models for speech recognition, the first learning data and the second learning data may include user speech data, a label value of the speech data, and user information corresponding to the speech data. Here, the input value may be data related to the acoustic model, and the label value may be data related to the language model, and the user information may be user identification information. Specifically, the input value may be data such as a waveform of the user's speech, and may include data such as intonation and pitch, and the label value may be text data in which the user's speech is converted into text. In addition, in order to identify learning data similar to the input value and label value of the second learning data from among the multiple learning data of the first external electronic device 200-1, the electronic device 100 may identify whether the waveform of the input value is similar, or whether the distribution of the label value represented in a vector format exists within a predetermined distance.

[0083] At the same time, the processor 130 can convert the voice data included in the first learning data by using a frequency filter. Specifically, the processor 130 can obtain a frequency filter that can generate voice data suitable for the second external electronic device 200-2 based on the feature information of the first external electronic device 200-1 and the feature information of the second external electronic device 200-2, and obtain the third learning data by using the obtained filter.

[0084] In addition, the processor 130 may perform the above operation when a predetermined time condition is satisfied. For example, the processor 130 may receive learning data from the first external electronic device 200-1 and the second external electronic device 200-2 at about dawn, and train the first artificial intelligence model and the second artificial intelligence model.

[0085] At the same time, according to another embodiment of the present disclosure, the first artificial intelligence model and the second artificial intelligence model may be artificial intelligence models for understanding natural language. In this case, the learning data may be information related to the intention and time slot of the user's voice included in the result of speech recognition obtained by the speech recognition model.

[0086] Figure 3 is a block diagram illustrating a detailed configuration of an electronic device according to an embodiment of the present disclosure.

[0087] Reference Figure 3 , in addition to the memory 110, the communication device 120, and the processor 130, the electronic device 100 may further include an input device 140, a display 150, and an audio output device 160. However, the components are not limited to the above components, and some components may be added or omitted as needed.

[0088] The input device 140 is a component for receiving the input of the user instruction. Here, the input device 140 may include a camera 141, a microphone 142, a touch panel 143, etc. The camera 141 is a component for acquiring image data around the electronic device 100. The camera 141 can take still images and moving images. For example, the camera 141 may include one or more image sensors (e.g., a front sensor or a back sensor), a lens, an image signal processor (ISP), or a flash (e.g., a light emitting diode (LED), a xenon lamp, etc.). The microphone 142 is a component for acquiring the sound around the electronic device 100. The microphone 142 can receive the input of an external sound signal and generate electronic voice information. In addition, the microphone 142 can use various noise elimination algorithms to eliminate the noise generated in the process of receiving the input of the external sound signal. The touch panel 143 is a component for receiving various user inputs. The touch panel 143 can receive the input of data through user operation. In addition, the touch panel 143 can be formed in combination with the display described below. Meanwhile, it is apparent that the input device 140 may include various components for receiving various data inputs in addition to the camera 141 , the microphone 142 , and the touch panel 143 .

[0089] The various components of the above-mentioned input device 140 can be used in various forms. For example, if the first external electronic device 200-1 includes a speech recognition model, and the second external electronic device 200-2 does not include a speech recognition model, and the second external electronic device 200-2 requires the result of speech recognition, the electronic device 100 can input the user voice input through the microphone 142 into the first artificial intelligence model received from the first external electronic device 200-1, and output the result value, and send the value to the second external electronic device 200-2. That is, the electronic device 100 can not only share learning data between the first external electronic device 200-1 and the second external electronic device 200-2, but also obtain the result value of the input value of the artificial intelligence model not included in the corresponding external electronic device on behalf of each external electronic device, and send the result value.

[0090] The display 150 is a component for outputting various images. The display 150 for providing various images can be implemented as a display panel in various forms. For example, the display panel can be implemented in various display technologies, such as a liquid crystal display (LCD), an organic light emitting diode (OLED), an active matrix organic light emitting diode (AM-OLED), a liquid crystal on silicon (LCOS), a digital light processing (DLP), etc. In addition, the display 150 can be combined with at least one of the front surface area, the side surface area, or the rear surface area of ​​the electronic device 100 in the form of a flexible display.

[0091] Specifically, the display 150 can display various setting information for sharing learning data between external electronic devices. That is, the sharing of learning data between external electronic devices can be performed automatically, but can also be performed by user instructions, and the display 150 can output various information for receiving user instruction input or querying user instructions.

[0092] The audio output device 160 is a component that outputs various types of notification sounds or voice messages and various types of audio data, and the audio processor performs various processing operations on these audio data, such as decoding or amplification, noise filtering, etc. The audio processor is a component that performs audio data processing. In the audio processor, various processing of audio data, such as decoding or amplification, noise filtering, etc., can be performed. The audio data processed in the audio processor can be output to the audio output device 160. In particular, the audio output device 160 can be implemented as a speaker, but this is only an example, and the audio output device 160 can be implemented as an output terminal capable of outputting audio data.

[0093] As described above, the processor 130 controls the overall operation of the electronic device 100. The processor 130 may include a RAM 131, a ROM 132, a main CPU 134, a graphic processor 133, first to nth interfaces 135-1 to 135-n, and a bus 136. Here, the RAM 131, the ROM 132, the main CPU 134, the graphic processor 133, and the first to nth interfaces 135-1 to 135-n may be connected to each other through the bus 136.

[0094] The ROM 132 stores a set of instructions for system booting, etc. When a power-on instruction is input and power is supplied, the main CPU 134 copies an operating system (O / S) stored in the memory 110 to the RAM 131 according to the instructions stored in the ROM 132, and boots the system by running the O / S. When the booting is completed, the main CPU 134 copies various types of application programs stored in the memory 110 to the RAM 131, and performs various types of operations by executing the application programs copied to the RAM 131.

[0095] Specifically, the main CPU 134 accesses the memory 110 and performs booting by using the O / S stored in the memory 110. Then, the main CPU 134 performs various operations by using various programs, contents, data, etc. stored in the memory 110.

[0096] The first to nth interfaces 135-1 to 135-n are connected to the above-mentioned various components. One of the interfaces may be a network interface connected to an external electronic device through a network.

[0097] In the following, reference will be made to Figures 4 to 8 Various embodiments according to the present disclosure are described.

[0098] Figure 4 and Figure 5 is a diagram illustrating a data sharing method according to various embodiments of the present disclosure.

[0099] Specifically, Figure 4 is a diagram showing a method of selecting learning data required for training the second artificial intelligence model in a case where the first external electronic device 200-1 is an external server storing a large amount of various learning data, and Figure 5 is a diagram illustrating a method of selecting learning data required for training a second artificial intelligence model in a case where the first external electronic device 200 - 1 is a personal terminal storing personalized learning data.

[0100] Reference Figure 4 , the first external electronic device 200-1 may include a first artificial intelligence model and a first database, and the second external electronic device 200-2 may include a second artificial intelligence model and a second database. Here, the first database and the second database may store a plurality of learning data.

[0101] First, the selection module can obtain first learning data similar to second learning data (including input value D2input and label value D2label) including input value and label value among the plurality of learning data in the second database from the plurality of learning data in the first database. Specifically, Figure 4 As shown, the selection module can compare at least one of the input value D2input and the label value D2label of the second learning data with at least one of the input value and the label value of each of the multiple learning data of the first database, and select similar data.

[0102] The selection module may use the first learning data including the input value D1 input and the label value D1 label similar to the second learning data as input values ​​to train the second artificial intelligence model.

[0103] Similar to the above method, the electronic device 100 may train the second artificial intelligence model by acquiring learning data similar to each of the plurality of learning data of the second database from the first database.

[0104] Reference Figure 5 , the electronic device 100 can use all of the multiple learning data in the first database as learning data for training the second artificial intelligence model. Figure 4Unlike the situation in which the learning data stored in the first external electronic device 200-1 and the second external electronic device 200-2 are all learning data received from the user, the electronic device 100 can train the second artificial intelligence model by using all of the multiple learning data of the first database.

[0105] Figure 6 is a diagram illustrating a method of converting learning data according to an embodiment of the present disclosure.

[0106] Reference Figure 6 , the selection module can obtain first learning data similar to second learning data (including input value D2input and label value D2label) including input value and label value in multiple learning data of the second database from multiple learning data of the first database. Specifically, Figure 4 As shown, the selection module can compare at least one of the input value D2input and the label value D2label of the second learning data with at least one of the input value and the label value of each of the multiple learning data of the first database, and select similar data.

[0107] The selection module can obtain the first learning data including the input value D1 input and the label value D1 label similar to the second learning data, and the conversion module can obtain the third learning data (input value D1' input and label value D1' label) based on the input value D1 input, the input value D2 input and the label value D1 label. That is, the learning data suitable for the second external electronic device 200-2 can be obtained by converting the first learning data as the third learning data.

[0108] As an embodiment for converting learning data, in a case where the first learning data and the second learning data are data for training an artificial intelligence model, the D1 input and the D2 input may be data related to a user voice received by each external electronic device.

[0109] Specifically, the electronic device 100 may add data of noise surrounding the first external electronic device to the second learning data (specifically, D2 input), and obtain third learning data (specifically, D1' input). For example, the electronic device 100 may obtain data of surrounding noise of the first external electronic device 200-1 according to the use environment of the first external electronic device 200-1, and add the obtained data of surrounding noise of the first external electronic device 200-1 and the D2 input to the time domain. As another example, the electronic device 100 may filter the D2 input in the frequency domain by using a frequency filter for the ambient noise environment of the first external electronic device 200-1. As another example, the electronic device 100 may obtain data of a non-speech part included in the D1 input by using voice activity detection (VAD), and add the data of the non-speech part and the D2 input to the time domain. Here, VAD is a technology for dividing the input user utterance into a part including the user's voice and a part not including the user's voice (a silent part or a non-voice part), and the electronic device 100 can obtain the third learning data by using the noise environment information included in the first external electronic device 200-1 using the non-voice part.

[0110] As another example, the electronic device 100 may acquire third learning data based on feature information of the first external electronic device 200 - 1 and the second external electronic device 200 - 2 .

[0111] Specifically, when the characteristic information of the first external electronic device 200-1 and the second external electronic device 200-2 is the specification information of the first external electronic device 200-1 and the second external electronic device 200-2, the electronic device 100 can consider the difference between the specification information of the first external electronic device 200-1 and the specification information of the second external electronic device 200-2 to obtain the third learning data.

[0112] Alternatively, in a case where the characteristic information of the first external electronic device 200-1 and the second external electronic device 200-2 is characteristic information related to noise input through the first external electronic device 200-1 and the second external electronic device 200-2, the electronic device 100 may obtain third learning data by using the difference between the noise input to the first external electronic device 200-1 and the noise input to the second external electronic device 200-2.

[0113] Alternatively, in a case where the characteristic information of the first external electronic device 200-1 and the second external electronic device 200-2 is the distance information between the position of the first external electronic device 200-1 and the position where the user voice is generated, and the distance information between the position of the second external electronic device 200-2 and the position where the user voice is generated, the electronic device 100 may acquire the third learning data based on the distance information between the position of the first external electronic device 200-1 and the position where the user voice is generated, and the distance information between the position of the second external electronic device 200-2 and the position where the user voice is generated. For example, it can be assumed that the user voice is input into the first external electronic device 200-1 within a short distance, and the user voice is input into the second external electronic device 200-2 from a long distance. Generally, there is a feature that, in the case where the user voice is input from a long distance, in the audio signal received by the device, the low frequency part is emphasized, and in the case where the user voice is input from a short distance, in the audio signal received by the device, the high frequency part is emphasized. Therefore, in order to convert the audio signal corresponding to the user voice received by the first external electronic device 200-1 into an audio signal suitable for the second external electronic device 200-2, signal processing such as reducing the high-frequency area of ​​the audio signal corresponding to the user voice received by the first external electronic device 200-1 and increasing the low-frequency part is required. Therefore, in view of the above, the electronic device 100 can convert the third learning data considering the distance information between the position of the first external electronic device 200-1 and the position where the user voice is generated and the distance information between the position of the second external electronic device 200-2 and the position where the user voice is generated. As described above, in order to convert the first learning data into the third learning data, the electronic device 100 can use a frequency filter.

[0114] At the same time, if Figure 5 As shown, the electronic device 100 can convert all of the plurality of learning data of the first database to adapt to the characteristic information of the second external electronic device 200-2. Figure 6 In the invention, an embodiment is described in which data similar to the second learning data is acquired and the acquired learning data is changed, but the present disclosure is not limited thereto, and the electronic device 100 can convert all the learning data of the first database without the process of acquiring learning data similar to the second learning data.

[0115] Figure 7 is a diagram for illustrating a method of acquiring learning data according to an embodiment of the present disclosure.

[0116] Reference Figure 7, the first external electronic device 200-1 is an electronic device that receives the user's learning data and is trained for a long time, and may include an artificial intelligence model that includes learning data related to multiple users and performs a large amount of learning. The second external electronic device 200-2 is, for example, an electronic device newly purchased by the user, and may include an artificial intelligence model that has a small amount of learning data of the user and has not yet been personalized.

[0117] In this case, if Figure 7 As shown, the electronic device 100 may input the D2 input received from the second external electronic device 200-2 as an input value into the first artificial intelligence model and output a result value. Here, when the D2 input is input into the first artificial intelligence model, the electronic device 100 may output a label value D1 label for the D2 input and a probability value Prob as to whether the D1 label is the label value of the D2 input. When the probability value Prob is equal to or greater than a predetermined value, the selection module may train the second artificial intelligence model based on the fourth learning data including the D2 input and the D1 label.

[0118] That is, in the case where the second external electronic device 200-2 is a newly purchased electronic device or an electronic device whose internal information has been initialized, etc., the label value and output value of the input data are incorrect. Therefore, the electronic device 100 can obtain the fourth learning data for the input value (D2 input) of the second learning data by using the first artificial intelligence model in which a large amount of learning has been performed.

[0119] Figure 8 is a diagram illustrating a method of sharing learning data according to an embodiment of the present disclosure. Fig. 9 is a diagram illustrating a method for a second external electronic device to convert learning data according to an embodiment of the present disclosure.

[0120] Reference Figure 8 and Fig. 9 In the case where the first artificial intelligence model and the second artificial intelligence model are artificial intelligence models for understanding natural language, the electronic device 100 may train the second artificial intelligence model by using first learning data having similar characteristics, and train the first artificial intelligence model by using the second learning data.

[0121] Specifically, the electronic device 100 may acquire the first learning data, and acquire the second learning data having the same or similar label as the label of the first learning data from the plurality of learning data in the second database. Alternatively, the electronic device 100 may obviously acquire the second learning data, and acquire the first learning data having the same or similar label as the second learning data from the plurality of learning data in the first database. Here, the labels in the artificial intelligence model for understanding language may be labels related to the intent and entity of the user's voice for the speech recognition result.

[0122] Meanwhile, in the above-mentioned various embodiments, the electronic device 100 is described based on an embodiment in which the electronic device 100 receives an artificial intelligence model and learning data of each of a plurality of external electronic devices from a plurality of external electronic devices and obtains learning data applicable to another external electronic device. However, the present disclosure is not limited to the above-mentioned embodiments, and the above-mentioned functions and operations of the electronic device 100 can be used for each of a plurality of external electronic devices.

[0123] Specifically, Fig. 9 As shown, the above-mentioned functions and operations of the electronic device 100 may be performed by the first external electronic device 200 - 1 or the second external electronic device 200 - 2 .

[0124] For example, the second external electronic device 200 - 2 may receive a plurality of learning data stored in the first external electronic device 200 - 1 and convert the received learning data into learning data suitable for an artificial intelligence model of the second external electronic device 200 - 2 .

[0125] Specifically, the second external electronic device 200-2 may receive a plurality of learning data stored in the first external electronic device 200-1, and identify first learning data corresponding to second learning data stored in the second external electronic device 200-2 among the plurality of learning data received from the first external electronic device 200-1.

[0126] In addition, the second external electronic device 200-2 may compare at least one of the input value and the label value included in the second learning data with the input value and the label value included in each of the plurality of learning data received from the first external electronic device 200-1, and identify the first learning data.

[0127] Then, the second external electronic device 200 - 2 may train a second artificial intelligence model used by the second external electronic device 200 - 2 based on the identified first learning data.

[0128] Specifically, the second external electronic device 200-2 can receive the feature information of the first external electronic device 200-1, and convert the first learning data into the third learning data based on the received feature information. At the same time, it is obvious that the same method as the various embodiments described above can be applied to the type and conversion method of the feature information. That is, the feature information may include at least one of the following items: feature information related to the voice input device included in the first external electronic device 200-1 and the second external electronic device 200-2, feature information related to the noise input through the voice input device of the first external electronic device 200-1 and the second external electronic device 200-2, and feature information related to the distance between the location where the user voice is generated and the first external electronic device 200-1 and the distance between the location where the user voice is generated and the second external electronic device 200-2.

[0129] Fig.10 is a flowchart illustrating a control method of an electronic device according to an embodiment of the present disclosure.

[0130] Reference Fig.10 In operation S1110, the electronic device 100 may receive, from each of the first external electronic device 200-1 and the second external electronic device 200-2, an artificial intelligence model used by each of the first external electronic device 200-1 and the second external electronic device 200-2 and a plurality of learning data stored in each of the first external electronic device 200-1 and the second external electronic device 200-2. In addition, the electronic device 100 may receive feature information of the first external electronic device 200-1 and the second external electronic device 200-2 and store the information.

[0131] In addition, in operation S1120, the electronic device 100 can identify first learning data corresponding to second learning data received from the second external electronic device 200-2 among multiple learning data received from the first external electronic device 200-1. Specifically, the electronic device 100 can compare the input value and label value of the first learning data with the input value and label value of the second learning data, and identify the first learning data.

[0132] Then, in operation S1130, the electronic device 100 may train the artificial intelligence model used by the second external electronic device 200-2 based on the acquired first learning data. Here, the electronic device 100 may train the second artificial intelligence model based on the third learning data, which is the learning data converted from the first learning data into a form suitable for the second external electronic device 200-2.

[0133] In operation S1140, the electronic device 100 may transmit the trained artificial intelligence model to the second external electronic device 200-2. Through the above process, the second external electronic device 200-2 may output a personalized result based on the second artificial intelligence model trained by various personalized learning data.

[0134] Meanwhile, the term "component" or "module" used in the present disclosure includes a unit composed of hardware, software or firmware, and it can be used interchangeably with terms such as logic, logic block, component or circuit. In addition, a "component" or "module" can be a component composed of an integrated body or a minimum unit or a part thereof that performs one or more functions. For example, a module can include an application specific integrated circuit (ASIC).

[0135] Various embodiments of the present disclosure may be implemented as software including instructions stored in a machine-readable storage medium, which may be read by a machine (e.g., a computer). The machine refers to a device that calls instructions stored in a storage medium and can operate according to the called instructions, and the device may include an electronic device (e.g., electronic device 100) according to the aforementioned embodiment. In the case where the instructions are executed by a processor, the processor itself may perform functions corresponding to the instructions, or may be executed by using other components under its control. Instructions may include code generated or executed by a compiler or interpreter. A machine-readable storage medium may be provided in the form of a non-transitory storage medium. Here, the term "non-transitory" only means that the storage medium does not include a signal and is tangible, but does not specify whether the data is stored in the storage medium semi-permanently or temporarily.

[0136] In addition, according to an embodiment of the present disclosure, the methods according to the various embodiments described in the present disclosure may be included in a computer program product for provision. A computer program product refers to a product that can be traded between a seller and a buyer. The computer program product may be distributed online in the form of a machine-readable storage medium (e.g., a compact disc read-only memory (CD-ROM)), or distributed through an application mall (e.g., an entertainment mall). In the case of online distribution, at least a portion of the computer program product may be at least temporarily stored in a storage medium, such as a manufacturer's server, an application mall server, and a memory of a relay server, or may be temporarily generated.

[0137] In addition, each of the components (e.g., modules or programs) according to the aforementioned various embodiments can be composed of a single object or multiple objects. In addition, in the above-mentioned corresponding subcomponents, some subcomponents can be omitted, or other subcomponents can be further included in various embodiments. Usually or additionally, some components (e.g., modules or programs) can be integrated into objects, and perform the functions performed by each component in the same or similar manner before integration. The operations performed by the modules, programs or other components according to various embodiments can be performed sequentially, in parallel, repeatedly or heuristically. Or, at least some operations can be performed or omitted in different orders, or other operations can be added.

[0138] While the present disclosure has been described with reference to various embodiments thereof, 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 by the appended claims and their equivalents.

Claims

1. A control method for an electronic device, the control method include: Receiving, from a first external electronic device and a second external electronic device, a first artificial intelligence model used by the first external electronic device and a second artificial intelligence model used by the second external electronic device, and a plurality of learning data stored in the first external electronic device and the second external electronic device, respectively; identifying, among the learning data received from the first external electronic device, first learning data corresponding to second learning data received from the second external electronic device; receiving second characteristic information of the second external electronic device; Based on the second characteristic information of the second external electronic device, converting the first learning data into third learning data suitable for training the second artificial intelligence model; training the second artificial intelligence model used by the second external electronic device based on the third learning data; as well as The trained second artificial intelligence model is sent to the second external electronic device.

2. The control method according to claim 1, further comprising: include: receiving first characteristic information of the first external electronic device, The converting of the first learning data into the third learning data includes: converting the first learning data into the third learning data based on the first characteristic information of the first external electronic device and the second characteristic information of the second external electronic device.

3. The control method according to claim 1, in, Identifying the first learning data includes: comparing at least one of an input value and a label value included in the second learning data with an input value and a label value included in the learning data received from the first external electronic device; and The first learning data is identified based on a result of the comparing.

4. The control method according to claim 2, in, The second artificial intelligence model includes an artificial intelligence model for speech recognition, and The multiple learning data include voice data, label values ​​of the voice data, and user information corresponding to the voice data.

5. The control method according to claim 4, in, Identifying the first learning data includes: comparing at least one of second voice data included in the second learning data, a second label value of the second voice data, and second user information corresponding to the second voice data with at least one of first voice data included in the learning data received from the first external electronic device, a first label value of the first voice data, and first user information corresponding to the first voice data; and The first learning data is identified based on a result of the comparing.

6. The control method according to claim 4, in, The first feature information and the second feature information include at least one of the following items: feature information related to a voice input device included in the first external electronic device and the second external electronic device, feature information related to noise input through the voice input device of the first external electronic device and the second external electronic device, and feature information related to a distance between a location where a user voice is generated and a location of the first external electronic device and the second external electronic device.

7. The control method according to claim 4, in, Converting the first learning data into the third learning data includes converting speech data included in the first learning data by using a frequency filter.

8. The control method according to claim 1, in, Receiving the first artificial intelligence model and the second artificial intelligence model includes: Based on a predetermined time condition, receiving the first artificial intelligence model used by the first external electronic device and the second artificial intelligence model used by the second external electronic device respectively; and receiving the plurality of learning data from the first external electronic device and the second external electronic device, the control method further comprising: receiving first feature information of the first external electronic device, and Converting the first learning data into the third learning data includes converting the first learning data into the third learning data based on the first feature information of the first external electronic device and the second feature information of the second external electronic device.

9. An electronic device, include: Memory; communication devices; as well as Processor, configured as: receiving, via the communication device, from a first external electronic device and a second external electronic device, a first artificial intelligence model used by the first external electronic device and a second artificial intelligence model used by the second external electronic device, and a plurality of learning data stored in the first external electronic device and the second external electronic device, respectively; identifying, among the learning data received from the first external electronic device, first learning data corresponding to second learning data received from the second external electronic device, receiving second characteristic information of the second external electronic device, Based on the second characteristic information of the second external electronic device, converting the first learning data into third learning data suitable for training the second artificial intelligence model, training the second artificial intelligence model used by the second external electronic device based on the third learning data, and The trained second artificial intelligence model is sent to the second external electronic device via the communication device.

10. The electronic device according to claim 9, in, The processor is further configured to: receiving first characteristic information of the first external electronic device via the communication device, and The first learning data is converted into the third learning data based on the first feature information of the first external electronic device and the second feature information of the second external electronic device.

11. The electronic device according to claim 9, in, The processor is further configured to: comparing at least one of an input value and a label value included in the second learning data with an input value and a label value included in the learning data received from the first external electronic device; as well as The first learning data is identified based on a result of the comparing.

12. The electronic device according to claim 10, in, The second artificial intelligence model includes an artificial intelligence model for speech recognition, and The multiple learning data include voice data, label values ​​of the voice data, and user information corresponding to the voice data.

13. The electronic device according to claim 12, in, The processor is further configured to: comparing at least one of second voice data included in the second learning data, a second label value of the second voice data, and second user information corresponding to the second voice data with at least one of first voice data included in the learning data received from the first external electronic device, a first label value of the first voice data, and first user information corresponding to the first voice data; as well as The first learning data is identified based on a result of the comparing.

14. The electronic device according to claim 12, in, The first feature information and the second feature information include at least one of the following items: feature information related to a voice input device included in the first external electronic device and the second external electronic device, feature information related to noise input through the voice input device of the first external electronic device and the second external electronic device, and feature information related to a distance between a location where a user voice is generated and a location of the first external electronic device and the second external electronic device.

15. An electronic device, include: Memory; communication devices; as well as Processor, configured as: receiving, from an external electronic device via the communication device, a plurality of learning data of a first artificial intelligence model stored in the external electronic device and suitable for use by the external electronic device, identifying, among the plurality of learning data received from the external electronic device, first learning data corresponding to second learning data stored in the electronic device, converting the first learning data into third learning data suitable for training a second artificial intelligence model used by the electronic device based on feature information of the electronic device, and A second artificial intelligence model used by the electronic device is trained based on the identified third learning data.

Images