158 results about "Ai systems" patented technology

Graphs are powerful structures made of nodes and edges. Information can be encoded in the nodes and edges themselves, as well as the connections between them. Graphs can be used to create manifolds which in turn can be used to efficiently train more robust AI systems. Systems and methods for graph-based AI training in accordance with embodiments of the invention are illustrated. In one embodiment, a graph interface system including a processor, and a memory configured to store a graph interface application, where the graph interface application directs the processor to obtain a set of training data, where the set of training data describes a plurality of scenarios, encode the set of training data into a first knowledge graph, generate a manifold based on the first knowledge graph, and train an AI model by traversing the manifold.

A telemedicine system with intelligent triage that replaces a remote doctor with a virtual doctor in some cases is disclosed. The telemedicine system may include: a triage procedure that employs a local attendant and software to determine if the patient should be seen by a human doctor or can be satisfactorily served by the virtual doctor AI system—the triage process is done at the start and throughout the consultation to ensure that the patient is handled appropriately, with several exit options for care to be provided by a human doctor if warranted; an AI system (e.g., an expert or knowledge-based system) comprising a knowledge-base, an inference engine, input module, output module, central / monitoring module, and learning module; and a remote control or operations center where the operation of the virtual doctors can be monitored by a set of human doctors who can choose to intervene if / when necessary.

The invention discloses an unmanned aerial vehicle routing inspection fault diagnosis system based on an integrated pod and an AI technology. The unmanned aerial vehicle routing inspection fault diagnosis system based on the integrated pod and the AI technology comprises an unmanned aerial vehicle, a deck, the integrated pod, a binocular vision module, an image and data transmitting module, an analysis control host and a fault diagnosis AI system loaded on the analysis control host, wherein the integrated pod is provided with an infrared detecting module, a visible light detecting module, a control module and an image and data transmitting module; the binocular vision module, the infrared detecting module, the visible light detecting module and the image and data transmitting module are connected with the control module; and the analyzing control host is wirelessly connected with the image and data transmitting module through an image and data receiving module. By the system, the laborintensity of power grid routing inspection maintenance workers is reduced, the working efficiency and working quality of monitoring of a power grid environment are improved, and costs for maintenanceare reduced; and real-time monitoring and multi-point monitoring to the power grid environment are implemented, and scruple of people to power frequency radiation during running of power grid equipment is eliminated.

A virtual assistant AI system that may be connected to a wide variety of user accounts such as financial accounts, social media, news, shopping, utilities and service providers, travel accounts, and other account types. The AI then continually monitors connected accounts for changes, analyzes changes when they occur and identifies any relationships or interactions between accounts and potential or actual implications of changes, and generates proactive notifications and provides them to the user who may then further investigate using virtual assistant AI clients.

The invention discloses a joint learning framework based on cooperation of a cloud server and IoT equipment. The method is characterized in that a cloud server and IoT equipment collaborative joint learning framework of a distributed AI system architecture is adopted, and the method specifically comprises the steps of cloud offline training, IoT equipment and cloud collaborative online joint training, IoT equipment and cloud collaborative reasoning and the like. Compared with the prior art, the joint learning framework has the advantages that the method adapts to different IoT use environmentsor use preferences of users, better performance effects of the local equipment model and the cloud model are realized, prediction accuracy and average reasoning execution time of the neural network in the IoT equipment are effectively improved, data privacy of local equipment users is protected, and the method is especially suitable for distributed application scenes with diversified data distribution.

The invention discloses an AI construction method for a hierarchical behavior tree. The AI construction method includes the steps that AI units in a highest layer priority determine an own strategy according to current conditions, and a corresponding instruction of the determined strategy is issued to AI units in a next-layer priority; AI units in a middle-layer priority execute corresponding operation according to the received instruction issued by the previous-layer priority and then issues a corresponding instruction to AI units in a next-layer priority, wherein the middle-layer priority includes one or more layers of priorities; all AI units in a bottommost-layer priority execute corresponding operation according to the instruction issued by the previous-layer priority, wherein each AI unit in any layer priority is constructed based on the behavior tree. The invention further discloses an AI construction device for the hierarchical behavior tree, a game AI construction method and device based on the behavior tree, a game control method and device and an AI system.

The invention discloses an intelligent simulation combat system for an air army unmanned aerial vehicle. The system comprises an unmanned aerial vehicle, the unmanned aerial vehicle carries an MAXSimsimulation platform, the MAXSim simulation platform comprises a general platform layer, a function module layer, a resource layer and a data management center, and the general platform layer comprisesan unmanned aerial vehicle intelligent autonomous combat system and a strategic weapon deduction system; wherein a scheme library comprises an intelligent scheme and a scenario scheme function, and the algorithm library comprises a damage algorithm and an evaluation algorithm function; the MAXSim simulation platform further comprises an expert AI system, a situation awareness subsystem, an evaluation screening subsystem and an intelligent scheme generation subsystem. According to the invention, battlefield situation analysis is carried out; through an AI expert system, based on the mutual influence of combat behavior models among different combat entities, an intelligent combat rule and an imaginary scheme are generated, and through deduction, analysis and evaluation of an intelligent autonomous combat process, the influence of intelligent autonomous combat research on novel equipment requirements, battlefield research requirements and future battlefields is displayed.

A reinforcement learning processor specifically configured to train reinforcement learning agents in the AI systems by the way of implementing an application-specific instruction set is disclosed. The application-specific instruction set incorporates ‘Single Instruction Multiple Agents (SIMA)’ instructions. SIMA type instructions are specifically designed to be implemented simultaneously on a plurality of reinforcement learning agents which interact with corresponding reinforcement learning environments. The SIMA type instructions are specifically configured to receive either a reinforcement learning agent ID or a reinforcement learning environment ID as the operand. The reinforcement learning processor is designed for parallelism in reinforcement learning operations. The reinforcement learning processor executing of a plurality of threads associated with an operation or task in parallel.

A method and a system are provided of testing AI applications / systems on a System for testing one or more artificial intelligence System (STAIS) connectable to under-test AI systems (UTAIS) via the AI test connectors / interfaces through AI test platform. The method includes test modeling, data preparation, testing script generation, testing automation, and quality assurance. The system automatically identifies, analyzes, and displays all quality issues of UTAIS.

The invention discloses an application program control method, device and equipment and a readable storage medium, and relates to the field of program design. The method comprises the following steps:receiving a service creation request sent by a program client; creating a service container according to the service creation request and obtaining running data of the target application program fromthe application program client; inputting the running data into a target operation matching model to obtain a target operation; and sending the target operation to the program client, and correspondingly controlling the operation of the target application program. Through presetting rules in a test server, when the target application program needs to be tested, the program client only needs to access the test server, the service container is created through the test server, the target operation is determined through the test server according to the operation data, an AI system does not need to be configured for the test process of the target application program in the program client for AI processing, and the AI processing efficiency is improved.

A system, method, computer program product, and a web site for evaluating a patent document by using an Artificial Intelligence (AI) have been provided. The patent document is characterized by patent indices arranged into a hierarchy, each index reflecting a different aspect of the patent document. The values of the patent indices are analyzed by an Artificial Intelligence (AI) system, which generates a verbal conclusion regarding the value of the patent document based on the performed analysis. A reason associated with the generated conclusion may also be generated. The AI system is a decision tree based AI system, which is implemented so as to be accessible via a network. The value of the patent document may be additionally visualized by using a color coding of the patent document based on values of its patent indices.

A body language system for determining a body language message of a living being in a context comprising an artificial intelligence system, said AI system running a computer program that: retrieves an image of said living being showing body language; labels said living being in said image, resulting in a labeled living being; determines said context from said image using a trained machine learning model; determines a baseline body language of said labeled living being from said image using a trained machine learning model; adapts a trained machine learning model of said AI system using said baseline body language and said context; applies the adapted trained machine learning model of said AI system to the one image for categorizing said body language resulting in a category, and applying said category for determining said body language message.

The invention discloses a development machine learning device, and the device comprises: a machine learning platform which is a platform for mining value information from mass data based on the FusionInsight HD distributed storage and a parallel computing technology; a deep learning platform which is an enterprise-level deep learning modeling platform and can enable customer algorithm developers to efficiently manage data sets, develop algorithm codes, evaluate models and predict service release experience and reduce deep learning modeling thresholds; and a reasoning platform which is mainly used for completing multi-algorithm unified management and task containerized heterogeneous resource unified scheduling, assisting clients to realize cluster computing power sharing and reducing the operation and maintenance cost of the AI system. The development machine learning device is a developer-oriented one-stop development platform, provides mass data preprocessing and semi-automatic labeling, large-scale distributed training, automatic model generation and end-edge-cloud model on-demand deployment capability, and helps a user to quickly create and deploy a model and manage full-cycle AI workflow.

The present invention is related to a computer-implemented method of training artificial intelligence (AI) systems or rather agents (Maximum Entropy Regularised multi-goal Reinforcement Learning), in particular, an AI system / agent for controlling a technical system. By constructing a prioritised sampling distribution q(ôg) with a higher entropy q(Ôg) than the distribution p(ôg) of goal state trajectories ôg and sampling the goal state trajectories ôg with the prioritised sampling distribution q(ôg) the AI system / agent is trained to achieve unseen goals by learning from diverse achieved goal states uniformly.

An Artificial Intelligence (AI)-based network control system includes an AI system configured to obtain data from a network having a plurality of network elements and to determine actions for network control through one or more Machine Learning (ML) algorithms; a controller configured to cause the actions in the network; and a safeguard module between the AI system and the controller, wherein the safeguard module is configured to one of allow, block, and modify the actions from the AI system.

The invention provides a shared handheld man-machine interaction device and a business mode thereof. An interactive platform, a central server, a cabinet and an interactive terminal are connected through a network. A family voice interaction connection is established for a user of the interaction terminal; by utilizing a service background AI system, a language semantic recognition system, biological feature recognition, virtual agent and an actuator system are constructed; voice semantic recognition and interactive voice response are completed; machine learning and decision management are perfected; interactive voice response questions of weather, traffic, clock, arithmetic, calendar, science, encyclopedic knowledge learning and the like are provided to a user; meanwhile, the real-time positioning data uploaded by the interaction terminal and the interaction platform can be received through the central server, and the data is compared with a set user age group, man-machine interactioncontent options and safety options of a central storage to judge whether to trigger an alarm or not; the handheld man-machine interaction device is light and handy and is convenient to carry; the hand-held use can be realized when people go out for activities; indoor voice control can be realized; the multifunctional voice interaction backpack can be put into a backpack, a luggage case, a file box and the like for monitoring important articles, can provide personal and property safety monitoring and alarm services for users, and can provide voice interaction responses in aspects of ancient poems, nursery rhymes, children's vocabularies, music, idioms, stories, English dialogue translation, life encyclopedia information, food recommendation and the like for the users.

According to an embodiment, there is provided a system and method for automatic AI controlled loop based song construction. It provides and benefits from a machine learning AI in a audio loop selection engine for the generation of a song structure and for the selection of fitting audio loops from a database of audio loops. In one embodiment, the instant method provides a music generation process that utilizes an AI system that has been trained and validated on a music item database to complete the creation of a music item given an incomplete song that was started but not finished by a user.

A patrol method based on a smart rural AI system is provided. There is no need to devote a large amount of police force to the search and arrest of criminals or suspects who have fled to the countryside. As long as the rural local police wear AR glasses to patrol, as long as the personnel they see can be uploaded to the township-level terminal server in real time with the rural scene, the township-level terminal server will automatically carry out face recognition for each personnel, so as to accurately identify the fugitive criminals or suspects. After the township-level terminal server identifies the fugitive criminal or the criminal suspect, it can also be automatically sent to a higher-level terminal server within the competent public security organ for police officers to give operational instructions to the local police in the countryside according to the actual situation in the countryside and the danger of the escaped criminal or suspect, to wait for assistance, to arrest directly or to follow up continuously in order to find the criminal dens, etc. The method effectively alleviated the current security problems in rural areas, but also enables the police efficiency of police officers to be qualitatively improved.

Provided is an artificial intelligence (AI) system for simulating human brain functions such as perception and judgment by using a machine learning algorithm such as deep learning, and an application thereof. Provided is a device and method for recommending products to a user on the basis of facial expression information of the user through the AI system.The method, performed by the device, of recommending products includes: displaying a product selected by the user; obtaining user's facial expression information with respect to the displayed product; determining the user's satisfaction with the displayed product based on the obtained user's facial expression information; selecting a product set to be recommended to the user from among a plurality of product sets based on the determined user's satisfaction; and displaying at least one product included in the selected product set.

In accordance with various embodiments, described herein is a system (Data Artificial Intelligence system, Data AI system), for use with a data integration or other computing environment, that leverages machine learning (ML, DataFlow Machine Learning, DFML), for use in managing a flow of data (dataflow, DF), and building complex dataflow software applications (dataflow applications, pipelines). Inaccordance with an embodiment, the system can provide data governance functionality such as, for example, provenance (where a particular data came from), lineage (how the data was acquired / processed), security (who was responsible for the data), classification (what is the data about), impact (how impactful is the data to a business), retention (how long should the data live), and validity (whether the data should be excluded / included for analysis / processing), for each slice of data pertinent to a particular snapshot in time; which can then be used in making lifecycle decisions and dataflow recommendations.

An AI (artificial intelligence) system has been developed that uses an AI module wherein the AI system captures, indexes, and extracts digital workflow of complex technical know-how for designing, manufacturing, operating, maintaining and servicing products, machines and equipment, and turns the digital workflow into a GPS-map like, step-by-step interactive workflow guidance. The AI system uses a workflow acquisition system, which captures and digitizes experts' knowledge and workflow as they are physically performing their work or task in a spatial environment. The workflow acquisition system includes one or multiple video input devices such as cameras that capture videos from multiple perspectives, including but not limited to side-view and point-of-view (POV) in which the cameras can be head-mounted, eye-wearable, or shoulder-mounted. The AI system and the AI module thereof analyzes and indexes the audios and every frame of the videos to extract the workflow content, such as objects, activities, and states, from the video using one or more AI methods, such as NLP (natural language processing) or computer vision, such as object detection and activity recognition. The extracted digital workflows, including step-by-step information, are stored preferably in a cloud based enterprise knowledge repository, which can be used to teach and train workers in these skilled trades and help speed up the learning curve for individuals learning a new skill such as those replacing more senior workers. Authorized users can access this digital workflow content as interactive how-to videos anytime, anywhere and learn at their own pace.

Imaging protocols are modified in real time through implementation of artificial intelligence. Key steps include: inputting of medical images and patient data; performing artificial intelligence analysis; outputting a potentially modified radiological imaging examination protocol; if applicable, an option for the radiologist to review the patient data, images acquired and the AI's potentially modified radiological imaging examination protocol; delivery of the modified radiological imaging examination protocol to the imaging device; and, if applicable, an option for the radiologist to provide feedback and re-train the artificial intelligence algorithm. Further, some of the images that are utilized by the AI system include processed images, such as segmented and filtered volume rendered images.

The invention discloses an intelligent AI system for traffic violation data processing. The intelligent AI system comprises an AI intelligent module, an alarm module, an image comparison module, a camera, a license plate recognition module, a traffic police violation processing platform and a communication module. The system has the beneficial effects that whether the vehicle driving into the intersection lane presses the line or not can be judged according to the corresponding early warning area set according to the lane distribution, so that the vehicle can normally drive at the intersection, the non-standard driving of the vehicle can be conveniently captured, and the vehicle information such as the license plate can be identified; and through the communication module, vehicle illegal driving in a short message mode is sent to a mobile phone of a vehicle owner, a driver can be warned in time to correct illegal driving behaviors, through the AI intelligent module, the number of times of illegal line pressing behaviors of vehicles of the same license plate is counted, and under a traffic monitoring platform, the frequency of monitoring the vehicles at an intersection road section can be increased.

The embodiment of the invention provides an Internet content pushing method based on big data demand analysis and an AI system. First user portrait feature distribution which does not cover user portrait features corresponding to one or more user community interaction activities of a target user in a plurality of preset user community interaction activities is sent to each content pushing service; each content pushing service executes initial interaction tracking aiming at the target user according to the first user portrait feature distribution, and performs further analysis based on an initial interaction tracking log after generating the initial interaction tracking log, so as to accurately obtain second user portrait feature distribution of the target user; and user portrait features related to community interaction are supplemented in time. Thus, the content pushing configuration rule can be transmitted to the target content pushing service based on the second user portrait feature distribution to request the target content pushing service to perform corresponding Internet content pushing on the target user, and the reliability of Internet content pushing is improved.

The invention discloses a marketing strategy effect evaluation method based on a retail AI system. The marketing strategy effect evaluation method comprises the following steps: A) loading related data conforming to a set format through a group of SQL; wherein the related data are behavior information and sales strategy release of historical customers; B) preprocessing the related data to obtain processed data; C) substituting the processed data into a multivariate linear regression algorithm for AI learning and training, solving a residual equation set of the processed data to obtain an optimal solution coefficient, and obtaining a strategy effect evaluation model according to the optimal solution coefficient; and D) inputting a simulation strategy or parameter by the user, substituting the simulation strategy or parameter into the strategy effect evaluation model, giving an estimated value and an estimated error range of a corresponding index by the retail AI system, and analyzing the degree of influence of each dimension on the index. According to the invention, a data-supported marketing strategy effect evaluation function is provided for the retail industry, and requirements on hardware and operation and maintenance personnel are low.

The embodiment of the invention discloses a hybrid intelligent social network system, a social method and device. The system comprises the hybrid intelligent nodes formed by mapping social entities, the social entities comprise the physical devices, the people / mechanisms or the network AI systems, a mesh connection relation is formed among the hybrid intelligent nodes, and the hybrid intelligent nodes receive and transmit messages through the social network according to a preset information routing strategy. The hybrid intelligent node comprises an AI control module and a human control module,and the AI control module determines whether an event is generated or not according to an autonomous operation result; and the human control module is used for receiving and processing a human / institution control instruction to form an event for the hybrid intelligent node to respond and process. According to the present invention, the hybrid intelligent social contact is formed among heterogeneous social entities, so that the social network is not only the cooperation among people, but also the cooperation among people, physical equipment and AI intelligent application programs can be realized, and the intelligent level of the society is improved.

The invention relates to a model updating method in the field of artificial intelligence (AI), and can provide a flexible and efficient optimization technology for a decentralized distributed AI system. The method can update the local model through the interaction of the working nodes in the neighborhood. According to the method, the model can be updated by receiving one or more kinds of information of the at least one parameter, the gradient and the impulse of the model transmitted by other working nodes in the neighborhood and combining the local model and the data.

An artificial intelligence (AI) system that utilizes a machine learning algorithm, such as deep learning, etc. and an application of the AI system is provided. A method, performed by a server, of integrating and managing a plurality of databases (DBs) includes obtaining a plurality of knowledge graphs related to DBs generated from the plurality of DBs having different structures from one another, inputting the plurality of knowledge graphs related to DBs into a learning model related to DB for determining a correlation between data in the plurality of DBs, and obtaining a virtual integrated knowledge graph output from the learning model related to DB and including information about a correlation extracted from the plurality of knowledge graphs related to DBs.

The embodiment of the invention provides an online e-commerce recommendation method based on big data and a big data AI system, and the method comprises the steps: obtaining a first business service attribute of a target e-commerce business and e-commerce big data of the target e-commerce business in a preset statistical segment, and based on this, according to the first business service attribute and the e-commerce big data, generating a target input component, and performing business service attribute updating on the target input component based on the business service attribute updating model to obtain a second business service attribute corresponding to the target e-commerce business. Therefore, in combination with the first business service attribute and the e-commerce big data of the target e-commerce business in the preset statistical segment, the initial business service attribute characteristics and the characteristics of the e-commerce big data can be considered, and the business service attribute is updated based on the characteristics, so the accuracy of subsequent e-commerce business recommendation is improved.

The embodiment of the invention provides a big data processing method based on smart medical security and a smart medical AI system. First session security evaluation value information and second session security evaluation value information of a medical access micro-service for session process scheduling at each session security path are predicted, so that each session security path can be considered from the two aspects of session security category probability dimension of entering the session security path and probability distribution dimension of occurrence of abnormal attack behaviors, further, abnormal data stealing and abnormal information guiding in the session process can be effectively avoided in the session switching process, and the security in the session switching process is improved.