29results about "Multimedia data querying" patented technology

The embodiment of the disclosure provides a method, apparatus, device, storage medium and a program product for information processing. The method comprises: obtaining, in response to receiving a user query from a user to a digital assistant from a user, auxiliary visual information acquired by a visual acquisition tool associated with the user based on the user query, an acquisition time of the auxiliary visual information being adjacent to an initiation time of the user query; and determining a reply corresponding to the user query based on the acquired auxiliary visual information and the user query with a machine learning model.

The application discloses a multi-source education data search method and device, equipment and a storage medium, and the method comprises the steps of: obtaining education information from a plurality of heterogeneous data sources through a hybrid search strategy, and performing recursive data association on the education information to generate an extended information set; performing data screening on the extended information set based on a pre-trained model to obtain a screened extended information set; receiving a user query request, using a multi-modal hybrid retrieval model to search information in the screened extended information set, and obtaining a search result. Compared with the prior art, the application eliminates information silos and improves the education information search matching accuracy by obtaining education information from a plurality of heterogeneous data sources through a hybrid search strategy, performing recursive data association on the education information to generate an extended information set, receiving a user query request, and using a multi-modal hybrid retrieval model to search information in the screened extended information set.

Some cases involve retrieval technology field, information retrieval method, device, electronic equipment, storage medium and program product are disclosed, the method comprises: obtaining first information;Obtain the first feature vector of the first information and the first discrete code word of the first information;Contrast the first discrete code word and the second discrete code word, obtain the first contrast result, the second discrete code word corresponds to the second information;Filter the second information by using the first contrast result, obtain the third information;Contrast the first feature vector and the second feature vector, obtain the second contrast result, the second feature vector corresponds to the third information;Filter the third information by using the second contrast result, obtain the fourth information, the fourth information is used to represent the retrieval result of the first information. Through the implementation of the information retrieval method, the retrieval accuracy and the overall retrieval efficiency can be improved.

Embodiments of the present application provide a multi-agent-based query request processing method and device and a service device, relating to the technical field of artificial intelligence. The multi-agent-based query request processing method comprises: decomposing an input query request to obtain at least one subtask and processing logic description information corresponding to each subtask; determining a task execution instruction for each subtask, the task execution instruction at least defining a search target, a constraint condition and a modal demand information of the subtask; based on the task execution instruction of each subtask, obtaining multi-modal reference information and structured reference information related to each subtask; and based on the multi-modal reference information and the structured reference information related to each subtask, obtaining response information of the query request. The present application provides sufficient and reliable support for language model response, thereby improving the accuracy and practicality of the system in processing complex query requests.

The present disclosure provides a network model training method, a data processing method, and an apparatus. The network model training method comprises: acquiring target sample data, wherein the target sample data comprises text sample data and image sample data; inputting the target sample data into a network model to be trained to obtain a sample recognition result; and adjusting a parameter of a text encoder on the basis of a text recognition result and first supervision data corresponding to the text recognition result, adjusting a parameter of an image encoder on the basis of an image recognition result and second supervision data corresponding to the image recognition result, and a hybrid image-text recognition result and third supervision data corresponding to the hybrid image-text recognition result, and adjusting a parameter of a hybrid encoder on the basis of the hybrid image-text recognition result and the third supervision data corresponding to the hybrid image-text recognition result to obtain the trained network model formed by the text encoder, the image encoder, and the hybrid encoder.

This application discloses a model processing method, a model processing device, a computer program product, and an equipment. The method includes: acquiring content data of multimedia resources and interaction data of sample objects with respect to the multimedia resources; adaptively training a recommendation model using the content data and interaction data to obtain a first recommendation model; acquiring collaborative features; the collaborative features include at least one of the following: multimedia collaborative features of multimedia resources and object collaborative features of sample objects, wherein the multimedia collaborative features represent the content representation of multimedia resources in the online recommendation scenario, and the object collaborative features represent the object representation of sample objects in the online recommendation scenario; optimizing and training the first recommendation model using the collaborative features to obtain a trained recommendation model; and using the trained recommendation model to recall multimedia resources recommended to objects. Using this application can improve the accuracy of recommending multimedia resources to objects.

The application relates to the technical field of special query processing, in particular to a data labeling distribution intelligent scheduling and permission management system, which comprises a behavior segmentation module, a modal disassembly module, a window mapping module and a query arrangement module, and an instruction generation module; the behavior segmentation module identifies discontinuous modes to construct switching paragraph information set; the modal disassembly module analyzes action type structure to generate operation path comparison set; the window mapping module reconstructs start and end boundaries to generate window section structure list; the query arrangement module filters matching paths to generate task query path calling table; and the instruction generation module constructs query scheduling instruction set. In the application, the operation interval discontinuous mode is identified, the image text voice synchronous position is located and the edge area is analyzed, the sound energy change point section is extracted, the segment boundary is reconstructed based on the action structure, the window section sequence is constructed, the query task field is mapped and the matching path is filtered, the task and the path are encoded and combined, the scheduling rigidity problem is solved, massive data efficient distribution and dynamic permission control are realized, and the task response speed and accuracy are improved.

Various embodiments relate to mobile surveillance systems. A system may include a mobile surveillance unit including at least one camera for capturing data including one or more objects. The mobile surveillance unit may further include an application program including an artificial intelligence (AI) model for generating first metadata for at least one object of the one or more objects. The mobile surveillance unit further includes a database for storing the first metadata and a communication device for receiving second metadata. The application program may be configured to: compare the second metadata to the first metadata; retrieve output data responsive to the second metadata matching the first metadata; and cause the output data to be conveyed to a remote device via the communication device. Associated methods are also disclosed.

A method and a computing device for expanding a scenario database is disclosed. The scenario database includes a plurality of existing scenario samples, each scenario sample including sensor data depicting a surrounding environment of a vehicle over a period of time, wherein each scenario sample is associated with a scenario embedding representing the scenario sample in a multi-dimensional space. The method includes obtaining an indication of an empty volume in the multi-dimensional space, the empty volume is a volume not covered by the plurality of existing scenario samples of the scenario database; transmitting a data collection request in view of the empty volume to one or more vehicles; receiving, from a vehicle of a fleet of vehicles, data indicative of a recorded scenario matching the data collection request; and storing the received data indicative of the recorded scenario to the plurality of existing scenario samples of the scenario database.

The present application relates to the technical field of information retrieval and credit leasing, and particularly relates to an intelligent retrieval system for high-concurrency credit leasing business data. The present application acquires multi-source data through a data acquisition unit, and after encapsulation by a data processing unit, an asynchronous queue dispatching vector generation unit calls an AI model to generate a feature vector. During retrieval, a retrieval recognition unit analyzes a retrieval request and determines a mixed retrieval weight, and performs a query by fusing semantic and keyword strategies. A management unit adaptively switches index types, adjusts retrieval nodes and index parameters based on the number of retrieval requests and concurrent pressure within a preset time. A calculation unit determines a retrieval representation value based on the response time of a standard query, an analysis unit determines a retrieval state according to the retrieval representation value, and dynamically adjusts the number of retrieval nodes based on the retrieval state, optimizes the preset number of retrieval requests and related parameters, and forms a closed loop of data-driven and autonomous optimization. The present application improves retrieval efficiency.

We provide the system. [Solution] A data receiving means for receiving user inquiries, A data analysis method that analyzes received inquiries to identify frequently occurring questions, Information generation means that generates the optimal answer based on the analyzed results, A multimedia generation means that integrates visual and audio information based on the generated answer, A data distribution means for providing generated multimedia content to users, An information processing system that includes this.

We provide the system. [Solution] A means of identifying spatial information based on location information, Means for generating and providing multilingual multimedia content related to space, A means of learning the user's behavior and interests and personalizing the content based on that, Means of providing information to users through visual and audio technologies, A means of analyzing and responding to conversations with users using dialogue technology, A system that includes this.

The present application discloses a model processing method and apparatus, a computer program product, and a device. The method comprises: acquiring content data of a multimedia resource, and interaction data between a sample object and the multimedia resource; using the content data and the interaction data to perform adaptive training on a recommendation model to obtain a first recommendation model; acquiring a collaborative feature, the collaborative feature including at least one of features extracted from the interaction data and representing correlations between objects, between objects and resources, and between resources; and using the collaborative feature to perform optimization training on the first recommendation model to obtain a trained recommendation model, the trained recommendation model being used for determining a multimedia resource recommended to an object.

The application provides a media information violation detection method and device, electronic equipment and storage medium, and relates to the technical field of violation detection. The method comprises the following steps: acquiring media information to be detected; determining key data in the media information, the key data being data used for evaluating whether a violation exists; reorganizing the key data based on a reorganization rule corresponding to an information type of the media information to obtain reorganized data; and matching the reorganized data with data in a preconfigured standard database to determine whether there is a violation in the media information; wherein the data recorded in the standard database is data representing that a violation exists or data representing that no violation exists. The violation detection method, device, electronic equipment and storage medium provided by the application can conveniently and quickly detect violations in various types of media information, and can meet the detection requirements of massive media information.

This application discloses a file processing method, an electronic device, and a storage medium. The file processing method includes: determining a first identifier vector for a file based on its position in a file set; generating a first storage location for storing the first identifier vector based on a hash function and the file's feature vector; determining a second identifier vector based on the hash table corresponding to the hash function and the first identifier vector; and storing the second identifier vector in a second storage location obtained by performing a pseudo-random permutation on the first storage location. The second storage location is used to locate the first storage location corresponding to the feature vector of the file to be searched when searching for a file. Thus, based on the second storage location and the position of the file to be searched in the file set, the corresponding file can be found without requiring the user to provide precise keywords for the search, thereby improving file search efficiency.

The herein disclosed technology relates to a method (100) for expanding a scenario database. The scenario database comprises a plurality of existing scenario samples, each scenario sample comprising sensor data depicting a surrounding environment of a vehicle over a period of time, wherein each scenario sample is associated with a scenario embedding representing the scenario sample in a multi-dimensional space. The method (100) comprises obtaining (S102) an indication of an empty volume in the multi-dimensional space, wherein the empty volume is a volume not covered by the plurality of existing scenario samples of the scenario database; transmitting (S106) a data collection request in view of the empty volume to one or more vehicles of a fleet of vehicles; receiving (S108), from a vehicle of the fleet of vehicles, data indicative of a recorded scenario matching the data collection request; and storing (S110) the received data indicative of the recorded scenario to the plurality of existing scenario samples of the scenario database. The disclosed technology further related to a computing device (200) thereof.

The technology disclosed herein relates to a method (100) for expanding a scene database. The scene database includes multiple existing scene samples, each scene sample including sensor data depicting the environment surrounding a vehicle over a period of time, wherein each scene sample is associated with a scene embedding representing the scene sample in a multidimensional space. The method (100) includes: obtaining (S102) an indication of an empty volume in the multidimensional space, wherein the empty volume is a volume not covered by the multiple existing scene samples of the scene database; sending (S106) a data collection request for the empty volume to one or more vehicles in a vehicle platoon; receiving (S108) data from the vehicles in the vehicle platoon indicating a recorded scene matching the data collection request; and storing (S110) the received data indicating the recorded scene into the multiple existing scene samples of the scene database. The disclosed technology further relates to its computing device (200).

The application discloses a document question and answer method and device and electronic equipment. The method comprises the following steps: performing similarity retrieval in a vector database according to a question vector to determine a target text paragraph, wherein the vector database comprises vectors of multiple text paragraphs; generating a text answer according to the question, the target text paragraph and a large language model; if the text paragraph further comprises a picture mark, querying corresponding target picture relationship data from a relationship database according to a picture identifier in the picture mark, wherein the relationship database comprises multiple picture relationship data, and the picture relationship data comprises a picture identifier, original picture data, a picture mark and picture position metadata; and embedding the original picture data into a corresponding position of the text answer according to the picture position metadata to obtain a result answer. The application improves the application effect in a rich text scene and improves the user experience.

The embodiments of the present disclosure provide a media file processing method, apparatus, device, readable storage medium and product, the method includes: receiving a material sent by a second user; providing a drawing material to the first user in response to a media file drawing command triggered by the first user in the receiving interface of the material, where the drawing material is a material sent by the second user to the first user; obtaining a material to be merged uploaded by the first user, determining a blending mode information, the blending mode information being selected by the first user or the blending mode information being preset based on the drawing material; performing a blending operation on the drawing material and the material to be merged according to the blending mode information to obtain a target media file; and posting the target media file. The creation of the target media file can be realized based on the material determined jointly by the first user and the second user, and further can meet the personalized needs of the user, improve the user experience, and improve the interest of creating media files.

A method and system for vector retrieval acceleration includes at least one calculation module and at least one sorting acceleration module. Each calculation module includes a first receiving end and a second receiving end for receiving query vector blocks and center vector blocks respectively, and performs operations on the received query vector blocks and center vector blocks, so as to obtain distance calculation results between each query vector and all center vectors, sorting acceleration module. The at least one calculation module corresponds one-to-one with the at least one sorting acceleration module. Each sorting acceleration module receives the distance calculation results output by the corresponding calculation module, sorts the distance calculation results, so as to obtain a sorting intermediate state, and then stores sorting intermediate states into a system memory.

In one or more implementations, a camera user interface of a client application may be accessed from a third-party application. The camera user interface may be modified based on input from a developer of the third-party application. Modifications to the camera user interface may be made according to a framework provided by a service provider related to the client application. The modifications to the camera user interface of the client application may be directed to modifying one or more user interface elements of an initial version of the camera user interface, modifying functionality available via the initial version of the camera user interface, or adding information related to the third-party application into the initial version of the camera user interface.

According to a local content providing method for social media proposed in the present invention, a local keyword may be received on social media, and content of a corresponding region and content of a surrounding region corresponding to the local keyword are provided at a preset ratio, so that even if a user does not know an exact region name, geography of the surrounding region, and the like, the user can receive and check even the content of the surrounding region of the searched region, thereby easily and quickly finding local information desired by the user.

This invention discloses a multimodal retrieval method combining image features and semantic understanding, belonging to the field of information retrieval technology. This application employs a design that combines image segmentation technology with semantic role labeling, fully mining the semantic information in image and text data. By generating semantically anchored visual codebooks and cross-modal anchored indexes, deep fusion of images and text is achieved. This method can more accurately correspond and align visual and semantic information in images and text, alleviating the problems of information loss and low retrieval accuracy caused by independent processing of image and text data in existing technologies, significantly improving the accuracy and efficiency of multimodal retrieval. Simultaneously, by using anchor vectors from both the image and text sides for indexing, visual codewords and semantic role slots are tightly integrated, establishing more accurate cross-modal matching for matching and indexing between different modalities.

A computer-accessible medium includes a multimedia-document integration module that includes a heterogeneous distinction identifier between a first multimedia-document and a second multimedia-document and includes an integrator of the first multimedia-document and the second multimedia-document into an integrated multimedia-document, a query module that includes a query-generator encapsulated in a corresponding number of enquiry / inquiry transmissions, the enquiry / inquiry transmissions being short-message-service text-messages, an acknowledgement / negative-acknowledgement transmission being a short-message-service text-message, and a variance-analytic module that includes a generator of quantitative variance from the acknowledgement / negative-acknowledgement transmission, the quantitative variance describing statistical variances and discrepancies within the first multimedia-document and within the second multimedia-document and between the first multimedia-document and the second multimedia-document.

A computer-implemented method includes receiving a query at a media streaming platform for media items that are to be provisioned by the media streaming platform. The method also includes analyzing the query to determine a level of breadth associated with the query, where the level of breadth indicates a degree of specificity associated with the query. The method further includes identifying media items that match the query then presenting the identified media items in an interactive user interface. The identified media items are visually presented in a manner that corresponds to the determined level of breadth associated with the query. Various other methods, systems, and computer-readable media are also disclosed.

An example operation includes one or more of obtaining, by a moving vehicle, a first data including one or more of at least one first video and at least one first image of a moving transport, analyzing the first data, by the moving vehicle, to determine an initial issue, querying, by the moving vehicle, a server for a second data based on the analyzing, wherein the second data is one or more of at least one second video and at least one second image of the moving transport at a previous time, verifying, by the moving vehicle, an issue exists based on a delta above a threshold between the first data and the second data, and sending, by the moving vehicle, the verified issue to the server.

A device, system, and non-transitory computer readable medium for providing a specialized agent comprising an application and a virtual assistant are disclosed. The virtual assistant includes a neural network trained with a machine learning algorithm to emulate a communication style of a character and wherein the virtual assistant is further trained on application metadata to a predict response to questions about the application.

The application discloses a kind of methods and devices for managing application, it is related to application development technical field.The specific embodiment of the method includes: the current parameter value of one or more characteristic parameters of multimedia file contained by application is optimized and analyzed, and the multimedia file is adjusted according to the result of analysis, and the application is kept alive by running the adjusted multimedia file.The embodiment of the application overcomes the problem that the keeping-alive effect of the application in the prior art is poor and resource consumption is high, and improves the user experience of using application push service.

We provide the system. [Solution] A means of storing profile data for collecting information based on user interests, A means of collecting data in real time from multiple sources and indexing it based on specific classifications, A means for filtering and prioritizing collected data by referring to stored profile data, A means for generating text and multimedia content based on filtered and prioritized data, A means of delivering the generated content to the user's device, A means of receiving feedback from users and updating the algorithm based on that feedback, A system that includes this.