426 results about "Multimedia information" patented technology

The invention discloses a video content semantic understanding and text description generation method based on deep learning, and relates to the technical field of multimedia information processing.The method comprises the steps that the semantic similarity of a text and a video frame is calculated through a CLIP model, related key frames are selected, and features are aggregated; respectively extracting audio, visual and semantic features; aligning different modal features by using self-attention, unifying dimensions of the LSTM, and then splicing and fusing; attention weights are calculated at a video level, a frame level and a channel level, and key information expression is enhanced; swin Transform encodes fusion features, and LSTM (Long Short Term Memory) decodes step by step to generate natural language description; and a text-video index database is constructed, and rapid retrieval is realized based on semantic similarity. According to the method, the mapping relation between the video features and the natural language is learned end to end through the deep learning model, dependence on a fixed template can be eliminated, and semantic description with various sentence patterns and coherent logic is generated.

The invention discloses an intelligent assistant system application interaction method based on a personal knowledge base, and relates to the technical field of large models, and the method comprises the steps: capturing multi-dimensional data of an application scene through multimedia information, so as to construct a more comprehensive prompt word engineering template; secondly, optimizing the cue word engineering template by accessing and utilizing personalized information closely related to the user in a pre-constructed personal knowledge base, and further performing scene optimization, entity optimization and multimedia recognition optimization, so that the cue words generated by the system can be ensured to better meet the actual requirements and context of the user; according to the method, more accurate and effective input is provided for subsequent large model processing, so that the intention of the user can be more accurately understood, diversified requirements of different users can be better met, response is made according to personalized information of the user, and the accuracy and efficiency of application interaction are improved.

The invention provides a conference information analysis method, electronic equipment and a storage medium. The method comprises the following steps: acquiring multimedia information in a conference process, wherein the multimedia information comprises audio information and / or video information; based on the multimedia information, conference information is determined, the conference information comprises a conference text and speaking times and speaking duration of each speaker participating in the conference, the conference text is composed of a plurality of text paragraphs arranged according to a time sequence, and each text paragraph of the plurality of text paragraphs corresponds to any speaker participating in the conference; based on the conference information, conference analysis data is determined, and the conference analysis data comprises at least one of the following data: discussion hotness, conference conclusion quality, conference keyword condition, conclusion contribution degree and conference efficiency; determining a conference quality score based on the conference analysis data; a conference summary is generated based on the conference information, the conference analysis data, and the conference quality score. According to the scheme, the conference content can be recorded and analyzed more comprehensively and accurately.

The application belongs to the technical field of multimedia information processing, and discloses a kind of monitoring video key frame extraction method, system, medium, equipment and terminal, collect original video stream data, original video stream data is decomposed into image frame set;Image frame set obtained by decomposition is sampled, and the image frame result set obtained by sampling is filtered;The image frame set after filtering is adaptively clustered, and the result after clustering is collected to form a video abstract.In order to better utilize the memory space of storage medium and let the user better quickly browse the general content of original video stream, the application provides a key frame screening algorithm for original video stream by sampling, filtering and clustering.The key frame extraction method of the application screens out similar frames, redundant frames and fuzzy frames in original video stream through key frame screening algorithm for original video data, forms a video abstract storage of a section of original video, thereby greatly reduces the occupied storage space.

The invention relates to the technical field of multimedia, in particular to a multimedia information display method and device, electronic equipment and a storage medium, and the method comprises the steps: displaying a personal homepage of a target account; when the personal homepage comprises a work published by the target account, displaying a thumbnail icon corresponding to a multimedia information set in a surrounding area of the work, the multimedia information set comprising multimedia information interacting with the target account; and in response to a trigger operation for the thumbnail icon, displaying the multimedia information in the multimedia information set. According to the method and the device, the operation path of checking the multimedia information interacted with the target account by the target account can be reduced, and the operation difficulty of checking the multimedia information interacted with the target account by the target account is reduced, so that the use experience of the target account is improved.

The invention provides a man-machine interaction method and system in an emergency scene. The method comprises the following steps: analyzing received multimedia information of a first-aid requester to obtain a first-aid situation perception model; according to the first-aid situation perception model, performing selection processing on a possible first-aid response strategy by using a deep reinforcement learning algorithm, and analyzing a relationship between a symptom and a possible cause of disease of the requester in combination with a medical knowledge map technology to generate a first-aid guidance scheme; based on the first-aid guidance scheme, performing real-time optimization processing on a rescue path through a traffic flow prediction algorithm, dynamically adjusting a driving route of a rescue vehicle in combination with road condition information collected by the Internet of Things technology, and generating an optimal rescue path and a latest rescue progress notification; and by using the optimal rescue path and the latest rescue progress notice, performing visualization processing on the first-aid operation process, and generating intuitive visual guidance. According to the technical scheme provided by the invention, the first-aid response speed, accuracy and effectiveness are remarkably improved.

The invention discloses a train fault state detection method and system based on artificial intelligence, and relates to the technical field of train fault state detection. The method comprises the steps that multiple types of sensors are arranged at key parts of a train, state data such as temperature, vibration, electricity, energy consumption and noise are collected in real time, local receiving, preprocessing and wireless uploading of the data are achieved by means of a mobile terminal, and a cloud platform conducts multi-dimensional analysis and fault recognition through an artificial intelligence algorithm on the basis of a health assessment model. According to the train fault diagnosis system and method, health indexes and early warning information are generated and pushed to operation and maintenance personnel in real time, the system supports overhaul process recording and multimedia information uploading, information archiving and analysis tracing are achieved, the accuracy and response efficiency of train fault diagnosis are remarkably improved, the operation and maintenance cost is reduced, and the safety and intelligent level of a railway system are enhanced.

An incomplete multi-view face image depth anchor image clustering method based on manifold relationship maintenance belongs to the field of face image data processing in multimedia information processing, and comprises the following steps: firstly, constructing a subspace feature for each view through a graph neural network, and then maintaining an original feature and the subspace features by adopting a manifold relationship; meanwhile, anchor points specific to the view angle are randomly selected to be combined with the common subspace, the similarity between the result and original data is compared, gradient descent optimization is carried out on the network for many times, spectral clustering is carried out after a common subspace anchor map with discrimination is obtained, and finally the clustering result and accuracy are obtained. Compared with other methods, the method is higher in accuracy and more stable in performance.

The invention discloses an audio deep forgery detection method and device, a terminal and a storage medium, and relates to the technical field of multimedia information security and artificial intelligence, and the method comprises the steps: constructing an audio deep forgery detection network model; according to the training set, performing decoupling stage training on an audio deep counterfeiting detection network model, and determining an initial training network model; performing meta-learning on the initial training network model according to the training set, and determining a target training network model; and obtaining a to-be-detected audio, inputting the to-be-detected audio into the target training network model, and determining an audio category corresponding to the to-be-detected audio. According to the method, meta-learning is adopted in the training stage of the audio deep forgery detection network model to force the audio deep forgery detection network model to learn general knowledge with higher generalization, so that the problem of poor detection performance caused by model overfitting in the face of an unknown vocoder and a natural scene in the prior art can be effectively solved.

The invention relates to the field of vehicle control, and provides a vehicle screen adjusting method and device, a vehicle, a first screen and a second screen, the first screen is at least used for displaying instrument information, and the second screen is at least used for displaying multimedia information and / or vehicle control information. Controlling the first screen image quality parameter to be adjusted to a first image quality parameter, and controlling the second screen image quality parameter to be adjusted to a second image quality parameter; receiving an image quality parameter adjustment instruction, and determining a target adjustment screen and a target image quality parameter; and comparing the target image quality parameter with the current image quality parameter of the target adjusting screen to obtain a first comparison result, and then adjusting the image quality parameter of the target adjusting screen. The linkage adjustment of the vehicle screen is cancelled, the corresponding image quality parameters are determined respectively, the requirement for reducing the energy consumption in the transportation mode is met, meanwhile, it is guaranteed that a driver clearly reads the basic information of the vehicle during transportation, and the driving safety is guaranteed.

Systems and methods for document processing that can process and understand the layout, text size, text style, and multimedia of a document can generate more accurate and informed document representations. The layout of a document paired with text size and style can indicate what portions of a document are possibly more important, and the understanding of that importance can help with understanding of the document. Systems and methods utilizing a hierarchical framework that processes the block-level and the document-level of a document can capitalize on these indicators to generate a better document representation.

The present invention relates to an intelligent operation and maintenance system based on an AI big model and RPA, which relates to the field of data processing. The system receives user input information in response to an RPA robot, performs semantic matching through an AI semantic model, and obtains consulting questions; processes the consulting questions through an answer material library, and matches a number of historical answers; performs centralized weight distribution on the several historical answers to obtain a number of centralized weights; processes the consulting questions through a question and answer log, and performs answer frequency weight distribution on the several historical answers to obtain a number of frequency weights; optimizes the several historical answers according to the several centralized weights and the several frequency weights, obtains a target answer, and feeds back the target answer to the RPA robot to reply to the user, thereby solving the technical problems that the intelligent operation and maintenance system is difficult to effectively handle complex and non-standardized user questions, the answer matching algorithm has low adaptability, and the multimedia information processing capability is limited.

Disclosed in the present application are an interaction method and apparatus, an electronic device and a readable medium. The method comprises: in response to a page display operation, displaying a target page, wherein the target page comprises a first region and a second region, the first region being used for displaying multimedia information that corresponds to a first account, and the second region being used for displaying attribute information of at least one recommended object that corresponds to a second account, and resource attribute information that corresponds to a candidate object set; and in response to an interaction operation for the candidate object set, performing preset interaction processing, which corresponds to the interaction operation, on a candidate object in the candidate object set.

The embodiment of the invention provides a shooting method and device, head-mounted equipment and a storage medium, and the method comprises the steps: obtaining to-be-shot information recognized by a camera shooting assembly after the head-mounted equipment connected with terminal equipment starts the camera shooting assembly; sending the to-be-shot information to the terminal equipment, so that the terminal equipment displays the to-be-shot information after receiving the to-be-shot information; and in response to a detected shooting instruction, generating multimedia information according to the to-be-shot information. According to the method, the head-mounted device can send the to-be-shot information to the terminal device for display in the shooting process of the head-mounted device, and a user can check the to-be-shot information through the terminal device when the head-mounted device does not have a display function, so that the user can check the shot content in time, and the user experience is improved. Whether the shot content is wanted by a user or not is known, so that the timeliness of displaying the information to be shot is improved; in addition, when the head-mounted device has the display function, the user can check the to-be-shot information through the terminal device without checking the to-be-shot information through the head-mounted device, and the cruising ability of the head-mounted device is improved.

The embodiment of the invention discloses a multimedia information pushing method and device, equipment and a medium. The method comprises the steps of obtaining current cargo associated data and a current user identifier of a to-be-pushed user, inputting the current cargo associated data into a trained real-time evaluation model, and outputting a current cargo evaluation result; processing the current cargo evaluation result according to a preset cargo classification dimension to obtain current cargo special data of the user to be pushed; determining whether a target user identifier which is the same as the current user identifier exists in preset historical user identifiers, and if the target user identifier exists, taking historical goods special data corresponding to the target user identifier as first reference goods special data according to a preset historical goods corresponding relation; and according to the first reference cargo special data and the current cargo special data, determining target multimedia information currently pushed to the to-be-pushed user. And the timeliness and the accuracy of the pushed target multimedia information are improved.

The invention discloses a telecommunication fraud early warning method. The system obtains telecommunication fraud case data through multiple channels of texts, voices, pictures, videos and links, and classifies the telecommunication fraud case data. For non-text data, the system converts the non-text data into a text through a technical means, and then carries out preprocessing, word frequency statistics and similarity calculation to generate training data. Meanwhile, the system obtains behavior data of the client from the application server, carries out telecommunication fraud detection, and sends out early warning and pushes a disposal instruction if a detection result exceeds a set threshold value. The system overcomes the defects that a traditional early warning mode depends on call surface data and is difficult to deeply analyze information content, enhances the recognition capability of multimedia information such as pictures and the like, and can effectively cope with various forms of telecommunication fraud. By combining the neural network and the cosine similarity algorithm, the system expands the early warning analysis range, and solves the problem that the rule is difficult to cover fraud numbers and information.

The invention relates to the technical field of multimedia information processing, and discloses a dynamic display adjusting method and system for multimedia interactive display, and the method comprises the steps: obtaining user real-time interaction data, user role attributes, content type preferences and historical interaction data, and generating real-time interest distribution points according to the user real-time interaction data, the user role attributes, the content type preferences and the historical interaction data; predicting a user behavior in combination with historical interaction data, and generating a pre-loaded content triggering condition; generating a preloaded content set by integrating the preloaded content triggering condition, the user role attribute and the content type preference; acquiring physical parameters of equipment, and constructing a coordinate matrix and a transformation matrix of cross-equipment content presentation; the content is distributed to the target device through the transformation matrix, and finally the display content is dynamically adjusted according to the real-time interaction data. The problem that in the multimedia interactive display technology, the content cannot be smoothly presented in real time during multi-screen switching is solved, seamless connection of the content during multi-screen switching is achieved, and real-time smoothness of multi-screen switching is ensured.

The present invention provides a control method and system for an exercise apparatus, and an exercise apparatus. The method comprises: acquiring an association relationship between multimedia information and working parameters of an exercise apparatus; and controlling the exercise apparatus on the basis of the multimedia information and the association relationship, or, on the basis of exercise data of a user within a preset time and the association relationship, determining target multimedia information to be played. The working parameters of the exercise apparatus can be automatically adjusted on the basis of the multimedia information, and there is no need for the user to adjust the working parameters of the exercise apparatus after determining the multimedia information; in addition, the played multimedia information can be adjusted on the basis of the working parameters of the exercise apparatus, and there is no need for the user to leave the exercise position to adjust the multimedia information, thereby simplifying the processes of the user adjusting the exercise apparatus and adjusting the played multimedia information during exercise, and improving the user experience of using the exercise apparatus during exercise.

The invention relates to an intelligent operation and maintenance system based on an AI large model and RPA, and relates to the field of data processing. User input information is received in response to an RPA robot, semantic matching is performed through an AI semantic model, and a consultation question is obtained; processing the consultation question through an answer material library, and matching a plurality of historical answers; performing centralized weight distribution on the plurality of historical answers to obtain a plurality of centralized weights; processing the consultation question through a question and answer log, and performing answer frequency weight distribution on the plurality of historical answers to obtain a plurality of frequency weights; the historical answers are optimized according to the concentrated weights and the frequency weights, target answers are obtained and fed back to the RPA robot to reply users, and the problems that an intelligent operation and maintenance system is difficult to effectively process complex and non-standardized user problems, the answer matching algorithm is low in adaptability, and the user experience is poor are solved. And the multimedia information processing capability is limited.

The invention provides an interactive template generation method and system based on webpage online arrangement in the technical field of multimedia information release, and the method comprises the steps: S1, enabling a server to arrange an interface template for multimedia information release online in a dragging interactive mode through a webpage template arrangement device, and packaging the interface template into a JSON data package; s2, after the client is started up, a login instruction is sent to the server through a preset communication protocol; s3, after the server verifies the login instruction, terminal information carried by the login information is registered into a pre-created database, and a JSON data packet is sent to the client; step S4, the client receives the JSON data packet in real time, and analyzes and renders the JSON data packet through WebVi ew so as to display the interface template; and step S5, the interface template automatically obtains the multimedia information based on the set template parameters, and fills and displays the multimedia information so as to complete the message sending operation. The method has the advantages that the convenience, the flexibility and the expansibility of signal sending are greatly improved.

The invention provides an information recommendation method and device, electronic equipment and a storage medium. The method comprises the steps that an information aggregation interface of a target object is displayed, and the information aggregation interface comprises at least one piece of multimedia information related to the target object; in response to a selection operation for the at least one piece of multimedia information, the selected multimedia information is played in a playing interface, and the playing interface comprises an interaction control; and in response to a trigger operation for the interaction control, displaying recommendation information and at least one interaction event with the target object. Through the method and the device, the attraction of the recommended information can be increased, so that the acceptability of the user for the recommended information and the user experience are improved.

The invention relates to the technical field of intention recognition, and discloses an intention recognition method and device based on a large model. The method comprises the following steps: acquiring user account information and multimedia information input by a user; wherein the multimedia information input types comprise texts, voices and images; wherein the user account information comprises a user name, a member grade and a registration date; performing feature extraction on the multimedia information to obtain a feature vector; calculating semantic correlation weights of the feature vectors, and sorting the semantic correlation weights to obtain labeled feature data; and selecting a corresponding preset decision tree model according to the user account information, and processing the labeled feature data by using the preset decision tree model to obtain a user intention recognition result. According to the method, the intention recognition accuracy and processing efficiency can be remarkably improved.

The invention provides an interaction method and device, equipment and a storage medium, and relates to the technical field of computers. The method comprises: in response to a preset trigger instruction, displaying multimedia information corresponding to a first media identifier in a media display area of a preset page, and displaying a target object set in an object display area of the preset page, the target object set being determined based on an object corresponding to a candidate media identifier in an identifier aggregation area, and in response to a switching instruction aiming at a candidate media identifier in the identifier aggregation area, determining the candidate media identifier aiming at the switching instruction as a second media identifier, and switching the multimedia information displayed in the media display area from the multimedia information corresponding to the first media identifier to the multimedia information corresponding to the second media identifier. By adopting the technical scheme, the information viewing convenience and the information switching efficiency can be improved.

The present disclosure provides a video-based interaction method, apparatus, and device, and a storage medium. The method comprises: displaying a video playback page; in response to a display trigger operation for a target function page acting on the video playback page, displaying the target function page; and displaying a preset video playback window on the target function page, and, on the basis of playback progress information of a first video, continuing to play the first video in the preset video playback window, the preset video playback window supporting switching and playback of a video on the basis of a multimedia information stream.

The invention discloses an interaction method and device, electronic equipment and a readable medium, in response to a page display operation, a target page is displayed, the target page comprises a first area and a second area, the first area is used for displaying multimedia information corresponding to a first account, and the second area is used for displaying multimedia information corresponding to a second account. The second area is used for displaying attribute information of at least one recommendation object corresponding to the second account and resource attribute information corresponding to the candidate object set; and in response to an interaction operation on the candidate object set, performing preset interaction processing corresponding to the interaction operation on the candidate objects in the candidate object set. In this way, the user can know the resource attribute information of the candidate object set of the currently selected recommendation object and perform interactive operation while watching the multimedia information, and the use experience of the user in the industry of instant retail is improved.

Embodiments of the present disclosure provide an interaction method and apparatus, an electronic device, a readable storage medium and a program product. The interaction method comprises: displaying a multimedia display page, wherein multimedia information displayed by the multimedia display page is first multimedia information corresponding to a target object; in response to an interaction operation for the multimedia display page, displaying an object aggregation page, wherein the object aggregation page comprises object attribute information corresponding to the target object and recommendation information corresponding to the target object, and the recommendation information is used for acquiring object attribute information of a recommended object corresponding to the target object; and in response to a trigger operation for the object attribute information corresponding to the target object or the recommendation information corresponding to the target object, displaying second multimedia information corresponding to the recommended object, wherein the second multimedia information corresponding to the recommended object comprises the object attribute information of the recommended object.

The present application relates to the technical field of computers. Disclosed are a network asset mapping method and apparatus, and a storage medium and an electronic device. In the method, a device receives at least one piece of response information sent by a target station on the basis of a network asset request, wherein the at least one piece of response information includes station service information of the target station and multimedia information borne on the target station, the multimedia information at least including text information and image information; and then the device extracts, from the text information, target entity information of a target entity to which the target station belongs, extracts, from the image information, external application information of an external application associated with the target station, and associates the extracted entity object information and external application information and the obtained station service information as network asset information of the target station, so as to obtain a corresponding target mapping result. By means of the present application, multi-dimensional mining of network asset information of a target station is realized, thereby improving the accuracy and comprehensiveness of network asset mapping.

The invention relates to the technical field of multimedia information processing, in particular to a recommended video feature storage and quick retrieval method based on picture similarity, which comprises the following steps: S1, establishing an alternative video service database; s2, original video feature extraction: after an original video is obtained, extracting picture feature value, character feature value and face feature value information of key frames of the original video by adopting a video key frame processing mode which is the same as that in the step S1; s3, retrieving characteristic values of alternative video picture frames; and S4, calculating the similarity of the alternative videos. According to the method, the accuracy and retrieval speed of video recommendation can be effectively improved, and more convenient and personalized video browsing experience is brought to users.

The invention provides a large-screen intelligent broadcast control system and an operation method thereof, and belongs to the technical field of multimedia information, the system comprises a content management subsystem, an intelligent play subsystem, a large-screen device and a user terminal device; the content management subsystem obtains the media content through interaction with the user terminal equipment, and audits, arranges, publishes and manages the media content; and the intelligent playing subsystem interacts with the large-screen equipment to realize communication, and controls the large-screen equipment and the played media content. According to the method, the content playing strategy of the large-screen equipment can be adjusted in real time according to actual requirements, the timeliness, accuracy and display effect of information release are improved, efficient management and intelligent playing of media content are achieved, the content safety and playing control efficiency of a system are improved, and finally the user experience is improved.

The invention relates to a digital exhibition hall server based on 3D scene editing and an application method of the digital exhibition hall server, a panoramic editor realizes a tool for migrating an exhibition industry from a traditional offline mode to an online mode, and helps a customer to build an online exhibition hall with characteristics or restore an offline exhibition hall in a one-to-one mode according to requirements to carry out online channel publicity. And meaningless multimedia information waste and marketing cost are fully reduced. By matching with the VR virtual roaming technology, audiences and visitors can browse exhibition halls all over the country without going out through the Internet, enterprise understanding and understanding of customers are enhanced, more useful information can be mastered through interaction, and higher influence and pertinence are achieved. The panorama editor can enable the user to preview the cloud exhibition hall online and provide sensory experience the same as that of entity exhibition. VR and other virtual reality online exhibition can be provided, a certain degree of interaction is provided, interactivity is provided, the attention of audiences is prevented from being distracted, and the audiences are concentrated on exhibition content.