Поиск патентных документов

BACKGROUND The present disclosure relates to an artificial intelligence (AI) apparatus and method for updating an artificial intelligence model, and more particularly, to an artificial intelligence apparatus and method for directly selecting training data to be used to train an artificial intelligence model and updating the artificial intelligence model. Recently, apparatuses for providing various functions using an artificial intelligence model generated using a machine learning algorithm or a deep learning algorithm are increasing. For example, apparatuses for interacting with users by speech using a speech recognition model or recognizing an object or a user using an image recognition model are increasing. Currently, due to limitations in storage space or computing power, most edge devices do not have artificial intelligence models installed therein.

BACKGROUND The present disclosure relates to an artificial intelligence (AI) apparatus and method for updating an artificial intelligence model, and more particularly, to an artificial intelligence apparatus and method for directly selecting training data to be used to train an artificial intelligence model and updating the artificial intelligence model. Recently, apparatuses for providing various functions using an artificial intelligence model generated using a machine learning algorithm or a deep learning algorithm are increasing. For example, apparatuses for interacting with users by speech using a speech recognition model or recognizing an object or a user using an image recognition model are increasing. Currently, due to limitations in storage space or computing power, most edge devices do not have artificial intelligence models installed therein.

TECHNICAL FIELD The present invention relates to an artificial intelligence server that can improve the performance of an artificial intelligence model by training the artificial intelligence model to be domain adaptation (domain adaptation) to the various domains that caused the incorrect answer. BACKGROUND ART Artificial intelligence is a field of computer science and information technology that studies a method for computers to do thinking, learning, and self-development that human intelligence can do and means enabling computers to imitate human intelligent behavior. In addition, artificial intelligence does not exist by itself, but is directly or indirectly related to other fields of computer science. Especially in modern days, artificial intelligence elements are introduced in various fields of information technology so that attempts are being actively made to solve problems in the field.

FIG. 5 is a diagram illustrating a method of operating an artificial intelligence system according to an embodiment of the present invention. The artificial intelligence system may include an artificial intelligence device 100-1, a first external artificial intelligence device 100-2, and a second external artificial intelligence device 100-3. Each of the artificial intelligence device 100-1, the first external artificial intelligence device 100-2, and the second external artificial intelligence device 100-3 may include all the components shown in FIG. 1 . Meanwhile, although the artificial intelligence system includes three artificial intelligence devices in FIG. 5 , this is merely an example and more artificial intelligence devices may be included. The microphone 122 of the artificial intelligence device 100-1 and the microphone 122 of the first external artificial intelligence device 100-2 receive a wake-up command (S501 and S503).

FIG. 5 is a diagram illustrating a method of operating an artificial intelligence system according to an embodiment of the present invention. The artificial intelligence system may include an artificial intelligence device 100-1, a first external artificial intelligence device 100-2, and a second external artificial intelligence device 100-3. Each of the artificial intelligence device 100-1, the first external artificial intelligence device 100-2, and the second external artificial intelligence device 100-3 may include all the components shown in FIG. 1 . Meanwhile, although the artificial intelligence system includes three artificial intelligence devices in FIG. 5 , this is merely an example and more artificial intelligence devices may be included. The microphone 122 of the artificial intelligence device 100-1 and the microphone 122 of the first external artificial intelligence device 100-2 receive a wake-up command (S501 and S503).

Referring to FIG. 9 , the electronic device of the first user may be a first artificial intelligence cooking device 700. Further, a processor of the first artificial intelligence cooking device 700 may capture the image 910 of the ingredients using a camera. For example, when the first artificial intelligence cooking device 700 is an oven, the camera of the first artificial intelligence cooking device 700 may capture a front of the first artificial intelligence cooking device 700 or a cooking chamber of the first artificial intelligence cooking device 700. The processor of the first artificial intelligence cooking device 700 may capture the image 910 of the ingredients and the final food product image 930 produced by the recipe 920, and transmit the captured images to the artificial intelligence server 100. Further, the processor of the first artificial intelligence cooking device 700 may perform the cooking based on the control information.

Referring to FIG. 9 , the electronic device of the first user may be a first artificial intelligence cooking device 700. Further, a processor of the first artificial intelligence cooking device 700 may capture the image 910 of the ingredients using a camera. For example, when the first artificial intelligence cooking device 700 is an oven, the camera of the first artificial intelligence cooking device 700 may capture a front of the first artificial intelligence cooking device 700 or a cooking chamber of the first artificial intelligence cooking device 700. The processor of the first artificial intelligence cooking device 700 may capture the image 910 of the ingredients and the final food product image 930 produced by the recipe 920, and transmit the captured images to the artificial intelligence server 100. Further, the processor of the first artificial intelligence cooking device 700 may perform the cooking based on the control information.

Meanwhile, the artificial intelligence execution acceleration system according to the exemplary embodiment is configured to have only an artificial intelligence execution acceleration server from which a system configuration of a learning step is removed, and the artificial intelligence execution acceleration server is configured to perform only an execution function of artificial intelligence. In the artificial intelligence execution acceleration system according to the exemplary embodiment, a learning function of artificial intelligence is removed to improve execution speed of functions of artificial intelligence in an edge terminal of an artificial intelligence system. In addition, by providing the artificial intelligence system having the learning function removed, the execution function of artificial intelligence may be accelerated and a memory required for artificial intelligence execution may be drastically reduced.

Meanwhile, the artificial intelligence execution acceleration system according to the exemplary embodiment is configured to have only an artificial intelligence execution acceleration server from which a system configuration of a learning step is removed, and the artificial intelligence execution acceleration server is configured to perform only an execution function of artificial intelligence. In the artificial intelligence execution acceleration system according to the exemplary embodiment, a learning function of artificial intelligence is removed to improve execution speed of functions of artificial intelligence in an edge terminal of an artificial intelligence system. In addition, by providing the artificial intelligence system having the learning function removed, the execution function of artificial intelligence may be accelerated and a memory required for artificial intelligence execution may be drastically reduced.

The artificial intelligence unit 130 of the artificial intelligence device 600 may acquire the command word. Specifically, the mic 122 may receive user voice, and the controller 180 may acquire a command word included in the received voice. In addition, the artificial intelligence unit 130 of the artificial intelligence device 600 may perform a function corresponding to the acquired command word. For example, when the acquired command word is “Please output a today's schedule”, the artificial intelligence unit 130 of the artificial intelligence device 600 may search for the today's schedule and output the result of search. In this case, the artificial intelligence unit 130 of the artificial intelligence device 600 may perform the function corresponding to the command word using the output unit included in the artificial intelligence device 600.

DESCRIPTION OF DRAWINGS FIG. 1 is a diagram showing the configuration of an artificial intelligence cleaner according to an embodiment of the present invention. FIG. 2 is a plan view of an artificial intelligence cleaner according to an embodiment of the present invention, and FIG. 3 is a bottom view of an artificial intelligence cleaner according to an embodiment of the present invention. FIG. 4 is a flowchart illustrating a method of operating an artificial intelligence cleaner according to an embodiment of the present invention. FIG. 5 is a view illustrating an example of determining whether a cleaning instruction image of an acquired image is recognized using a cleaning instruction image recognition model according to an embodiment of the present invention.

... <Artificial Intelligence (AI)> Artificial intelligence refers to the field of studying artificial intelligence or methodology for making artificial intelligence, and machine learning refers to the field of defining various issues dealt with in the field of artificial intelligence and studying methodology for solving the various issues. Machine learning is defined as an algorithm that enhances the performance of a certain task through a steady experience with the certain task. An artificial neural network (ANN) is a model used in machine learning and may mean a whole model of problem-solving ability which is composed of artificial neurons (nodes) that form a network by synaptic connections. The artificial neural network can be defined by a connection pattern between neurons in different layers, a learning process for updating model parameters, and an activation function for generating an output value.

... <Artificial Intelligence (AI)> Artificial intelligence refers to the field of studying artificial intelligence or methodology for making artificial intelligence, and machine learning refers to the field of defining various issues dealt with in the field of artificial intelligence and studying methodology for solving the various issues. Machine learning is defined as an algorithm that enhances the performance of a certain task through a steady experience with the certain task. An artificial neural network (ANN) is a model used in machine learning and may mean a whole model of problem-solving ability which is composed of artificial neurons (nodes) that form a network by synaptic connections. The artificial neural network can be defined by a connection pattern between neurons in different layers, a learning process for updating model parameters, and an activation function for generating an output value.

In such a place as school, orphanage, or daycare center, the artificial intelligence object of the artificial intelligence transparent display may communicate with infants or students and provide to a curriculum. Besides, the artificial intelligence object of the artificial intelligence transparent display may serve as a conversation partner for elderly people or provide programs in a welfare facility or may be utilized for other various purposes. The artificial intelligence object may be downloaded by the user's selection. For example, the user may download the artificial intelligence object from a server for download, e.g., App store server, and the downloaded artificial intelligence object may be output on the artificial intelligence transparent display. As the artificial intelligence object, various characters may be developed, and sale of such artificial intelligence objects may be a revenue model.

The method for operating the artificial intelligence system according to the first embodiment of the present disclosure includes receiving, by a specific electronic device 300, input data and transmitting the received input data to the artificial intelligence server 100 (S605), receiving, by the artificial intelligence server 100, the input data from the specific electronic device (S610), inputting, by the artificial intelligence server 100, the input data to a common model to obtain an intermediate result value (S615), applying, by the artificial intelligence server 100, personalized information corresponding to the specific electronic device to a personalized model (S620), inputting, by the artificial intelligence server 100, the intermediate result value into the personalized model to obtain a final result value (S625), transmitting, by the artificial intelligence server 100, the final result value to the specific electronic device 300 (S630), and receiving, by the artificial intelligence ...

Embodiments of the invention moreover relates to a system for transforming a trained artificial intelligence model into a trustworthy artificial intelligence model, comprising: a user interface component to enable provision of the trained artificial intelligence model, a memory storing the trained artificial intelligence model and user assignment information, a computing component for generating generic samples based on a validation data set, wherein the validation data set is determined based on training data of the trained artificial intelligence model, and for transforming the trained artificial intelligence model by optimizing a calibration based on the generic samples. For example, the computing component may comprise a central processing unit (CPU) and a memory operatively connected to the CPU.

In step 216, updated self-learning software for execution of the communications between the master real-time artificial intelligence module and the plurality of real-time artificial intelligence modules is generated in response to the updated models. The updated self-learning software is executed. In step 218, a query associated with a Web search is executed in response to user input with respect to executing the updated self-learning software. Executing the query may include: 1. Parsing semantics associated with indexing functionality of the master real-time artificial intelligence module and the plurality of real-time artificial intelligence modules. 2. Parsing code based language of the master real-time artificial intelligence module and the plurality of real-time artificial intelligence modules. 3. Executing a real time query of the master real-time artificial intelligence module. 4.

SUMMARY An object of the present invention is to an artificial intelligence electronic device capable of determining an artificial intelligence electronic device which will perform a function when a plurality of artificial intelligence electronic devices receives speech input of a user.

Meanwhile, the processor may perform the object recognition with respect to an object by using the parameter of the trained artificial intelligence model. In one embodiment, the memory may store entire parameters of the artificial intelligence model. The entire parameters of the artificial intelligence model may be initial parameters (parameters of the artificial intelligence model set before the artificial intelligence model is mounted on the artificial intelligence moving agent). In addition, when the training of the artificial intelligence model is completed, a part of the entire parameters may be changed into new parameters. In this case, the processor may newly store the entire parameters, which are partially changed into the new parameters, in the memory, and the object recognition may be performed by using an artificial intelligence model set with the newly stored entire parameters.

The computer-implemented method of clause 1, wherein determining that the content is content generated by an artificial intelligence source comprises accessing an artificial intelligence source list comprising a listing of artificial intelligence sources, determining the source of the content, comparing the source of the content against the artificial intelligence source list, and determining that the source of the content is from one of the artificial intelligence sources. Clause 3.