450results about "Multimedia data clustering/classification" patented technology

A method, computer program product and computing device for generating a first user profile associated with a first user of a media distribution system. The first user profile being stored on a first personal media device associated with the first user. The first user profile including at least a user identifier for identifying the first user within the media distribution system. Communication is established with a second personal media device associated with a second user of the media distribution system. The first user profile is transmitted to the second personal media device.

An information processing apparatus selects a proper content that well matches preference of a user and recommends it. A matrix calculator acquires M (one or more) feature vectors CCV whose elements are given by weight values assigned to a total of N (two or more) pieces of content meta information and context information. The matrix calculator produces a matrix CCM whose columns are given by the M feature vectors CCV and converts it into an approximate matrix CCM* by modifying the weight values of the respective elements of the M feature vectors CCV such that correlations of elements among the M feature vectors CCV are emphasized. Based on the approximate matrix CCM*, a user preference vector (UPV) generator produces a user preference vector UPV*. A matching unit calculates similarity between the user preference vector UPV* and a feature vector CCV produced from new content meta information or context information.

Exemplary embodiments of the present invention comprise a tool kit for the design and execution of research protocols, for example, comprising a web-based research protocol composer for generating a data processing protocol; a collector for collecting data entered via an internet based questionnaire; a database for data; and a protocol processing routine for processing data and generating an output. For example, the toolkit suitably serves as a “hub” of a global research environment (e.g., connected via the Internet or other network) that creates and connects communities of researchers and facilitates the sharing of software and results. Additionally, the present invention provides standardized and custom online “programs” or “scripts” (referred to herein as “protocols”) that define and guide events in experiments, clinical trials, surveys, and other kinds of studies. These protocols are created and executed online.

A system for imaging a scene. The system includes a plurality of cameras, each camera including an image sensor and a field steering mirror. The system also includes a controller and a storage device. The controller is coupled to the cameras and the storage device and is configured to direct the field steering mirrors of the cameras to collect a plurality of image tiles of a scene. The is also configured to store the image tiles in the storage device with time stamps and location information.

The invention provides a cross-modal generalized zero sample retrieval method based on a dual learning generative adversarial network. The method comprises: constructing a generative adversarial network based on dual learning; mapping the high-dimensional visual features of different modes to a common low-dimensional semantic embedding space; secondly, constructing multiple constraint mechanisms to perform cyclic consistency constraint, generative adversarial constraint and classifier constraint so as to maintain visual-semantic consistency and generated feature-source feature consistency, andperforming cross-modal retrieval after training of the whole network, so that the model is more powerful in performance in generalization of zero-sample retrieval. Meanwhile, in the whole training process, paired multimedia data pairs on the pixel level do not need to serve as training samples, only paired data on the category are needed, so that the complexity and expensive cost of data set collection are reduced, the retrieval effect is better, and performance improvement is more obvious in the zero-sample generalization retrieval problem.

A method is provided for simultaneously recognizing facial attributes and identity to organize photo and / or video albums, based on modifying an efficient convolutional neural network (CNN) which extracts facial representations suitable for face identification and attribute (age, gender, ethnicity, emotion, etc.) recognition tasks. The method enables to process all the tasks simultaneously, without a need for additional CNNs. As a result, a very fast facial analytic system is provided, and the system can be installed onto mobile devices.

The invention provides a multimedia data processing method and device, equipment and a storage medium. The method comprises the steps of obtaining a plurality of target frame images in a to-be-processed video file to obtain a target frame image sequence; segmenting the target frame image sequence into a plurality of video clips, and determining video features of each video clip; determining candidate video clips from the multiple video clips based on the video features of the video clips, wherein the candidate video clips and reference video clips in a video feature library meet a first similarity condition; and when determining that the candidate video clip and the reference video clip meet a second similar condition based on the image features of each target frame image contained in thecandidate video clip, determining that the reference video clip and the candidate video clip are similar video clips. By means of the method and device, the calculation speed of similarity calculationcan be increased, and the accuracy can be guaranteed.

The invention discloses a cross-modal retrieval method based on a graph convolutional neural network. The cross-modal retrieval method comprises four processes of network construction, data set preprocessing, network training and retrieval and precision testing. Semantic representations in an image mode and a text mode are respectively learned by using a graph convolutional neural network; the cross-modal retrieval method can help to process the potential relationship among modal features, introduces the associated data of the third modal into the cross-modal retrieval method to reduce the semantic gap among the modals, and can significantly improve the accuracy and stability of cross-modal retrieval, thereby realizing accurate cross-modal retrieval.

The invention discloses a CMR model for uniformly retrieving cross-media information, and aims to provide a cross-media retrieval model with accurate and rapid information. The method is realized through the following technical scheme: a multi-modal media information semantic feature unified expression and association module queries heterogeneous information input by the input and cross-media datamodule; the multi-modal semantic features are mapped into the same feature space and a multi-modal semantic association rule is constructed based on the multi-modal semantic feature extraction resultand the mapping from the bottom-layer features to the high-layer semantic features, thereby realizing the association between the bottom-layer features and the high-layer semantic features of the cross-media information and the association of the high-layer semantics of different modal information; the cross-media data index construction module is used for establishing a multi-dimensional retrieval index for the multi-modal data characteristics; the cross-media retrieval model construction module realizes unified retrieval of multi-modal information based on ontology, semantic network and knowledge graph technologies; and conflict detection and self-organization of retrieval results are realized through a multi-modal retrieval result association verification and organization module.

The invention discloses a sensitive data discovery method based on text recognition. The sensitive data discovery method comprises the following steps of S01, extracting the sample data; S02, constructing a training sample, collecting a text data set, and constructing the training sample; S03, training a sample annotation model, obtaining a training sample based on S02, and training a text annotation model; S04, constructing data features; S05, constructing a training set, carrying out label description on the data set obtained in the S04 to form a training set for constructing a classification judgment model; S06, constructing a classification judgment model, and forming a variable prediction model according to the training set obtained in the S05; S07, testing the model. Through the identification of the data variables, the sensitive data can be accurately and efficiently judged and identified under the condition that the data dictionary and the matching rules are incomplete, and theconsistency of identification and classification results is ensured.

The embodiment of the invention provides a Tibetan language-based multi-modal emotion calculation method and system, and a server. The method comprises: firstly, obtaining Tibetan language data to beclassified; collecting video signals, voice signals and text information from the Tibetan language data; then, extracting high-level video features, high-level voice features and text features in a classification emotion corpus, respectively extracting the high-level video features, the high-level voice features and the text features, performing learning based on a deep learning model to obtain high-level fusion features, and finally, classifying the high-level fusion features in the classification emotion corpus based on SVM and storing the high-level fusion features in the classification emotion corpus. Therefore, the blank state of the Tibetan language in sentiment analysis can be filled. A basic corpus is provided for Tibetan multi-modal sentiment analysis. The Tibetan language data sentiment recognition method based on the three modes is beneficial to development of Tibetan language multi-mode sentiment analysis, the natural language processing capacity and the intelligent sentiment recognition capacity of the Tibetan language can be promoted, the artificial intelligence information processing capacity of the Tibetan language is improved, and in addition, the sentiment recognition rate of the Tibetan language data can be effectively increased under the condition of mutual fusion of the three modes.

A method for automatically providing key moments in a multimedia content on an electronic device and an electronic device therefor are provided. The method includes determining a navigation behavior of each user of the multimedia content during playback, determining a plurality of key moments in the multimedia content based on the navigational behavior, the plurality of key moments including a positive key moment, a negative key moment, and a neutral key moment, storing the plurality of key moments, detecting a playback event of the multimedia content by a candidate user, retrieving at least one key moment from the plurality of key moments in the multimedia content based on the candidate user, and displaying an actionable user interface including the at least one key moment.

The invention provides a few-sample cross-modal hash retrieval common representation learning method. According to the method, a oneself knowing-adversary knowing network is designed, mainly relates to two modules: a oneself knowing module and an adversary knowing module. The oneself knowing module can fully utilize the information hidden in the data itself, fuse the features of different levels,and extract more global features; on the basis of the oneself knowing modules, the adversary knowing module carries out modeling on the correlation of all the samples, and the nonlinear dependence relationship between the data is captured, so that the common representation of different modal data can be better learned. And finally, a loss function for maintaining intra-modal and inter-modal similarity is established, and training optimization is carried out on the network. According to the method, the problem of data imbalance under the condition of few samples can be effectively solved, and more representative common representation can be learned, so that the cross-modal retrieval precision is greatly improved.

There is disclosed a system and method for processing multi-modal collaboration. In an embodiment communication received from multiple modes are converted into a common format. Using various conversion modules, the communication may be converted into a common electronic text format (e.g. ASCII text) that contains keywords. Once the communication is converted into a common format, the information contained in the communication may be analyzed and consolidated into related areas or topics. The consolidated information may then be searched for common references in order to augment the information context.

The invention provides an error book management method and device. The method comprises the following steps: shooting homework and transmitting the homework to a background; performing search matchingwith a background tagged test question database; converting paper test questions into standard fonts and displaying the standard fonts on the flat plate; matching subjects, grades and knowledge pointinformation of the test questions by the photographed test questions according to the tagged information of the test question bank; carrying out wrong question review: synchronously presenting the analysis answers, the video explanations and the variant exercises of the knowledge points of the test questions; carrying out difficult problem printing: linking and printing all associated informationby generating a two-dimensional code of the test question; carrying out code scanning learning. A paper wrong question paper is printed by scanning a two-dimensional code, a student answers and reviews questions according to a traditional learning mode, and meanwhile, each question can be explained by scanning the two-dimensional code to see a video, and one-back-three practice service is provided. The pains that children do not do homework and parents do not teach are solved, and the burden of the parents for guiding children to do homework is reduced.

Disclosed embodiments provide a multimedia file categorizing, information processing, and model training method, system, and device. The method comprises determining a plurality of feature combinations according to respective feature sets corresponding to at least two types of modality information in a multimedia file; determining a semantically relevant feature combination by using a first computational model according to the plurality of feature combinations; and categorizing the multimedia file by using the first computational model with reference to the semantically relevant feature combination. The technical solutions provided by the disclosed embodiments identify a semantically relevant feature combination in a process of categorizing a multimedia file by synthesizing features corresponding to a plurality of modalities of the multimedia file. The semantically relevant feature combination has a stronger expression capability and higher value. Categorizing the multimedia file by using this feature combination can effectively improve the categorization accuracy of the multimedia file.

A method, computer program product and computing device for presenting identifying information associated with at least one media content item, the identifying information being presented to a user on a personal media device. A purchase request to purchase the at least one media content item associated with the identifying information is received. Upon receiving the purchase request, a media content item identifier associated with the at least one media content item is stored in a purchase queue to initiate purchase of the at least one media content item at a later time.

The embodiment of the invention provides a 5G financial message data processing method, a financial institution and an operator service device, which can be used in the fields of 5G technology and financial technology; the method comprises the following steps: receiving an interaction message sent by an operator; recognizing key words and material identifications in the self-interaction message, and selecting the material resource obtaining addresses from an address base, wherein the address base is used for storing the corresponding relation between all the material identifications and all the material resource obtaining addresses obtained from an operator in advance; and sending the material resource acquisition address to an operator, so that the operator extracts a corresponding financial interaction material from a preset material library based on the acquired material resource acquisition address, and sending a 5G financial message containing the financial interaction material to 5G client equipment. According to the method and the device, the convenience and the intelligent degree of short message interaction between the user and the financial institution can be effectively improved, so that the user can more conveniently and efficiently enjoy financial services provided by the financial institution.

The invention discloses a multimedia resource classification method, a device, a computer device and a storage medium, belonging to the computer technical field. The method comprises the following steps of: acquiring multimedia resources and extracting a plurality of characteristic information of the multimedia resources; Clustering a plurality of feature information to obtain at least one clustering set, determining clustering description information of each clustering set, each clustering set comprising at least one feature information, and each clustering description information for indicating a feature of a clustering set; Determining at least one target feature description information of the multimedia resource based on the clustering description information of each clustering set, each target feature description information representing an association between one clustering description information and the rest of the clustering description information; The multimedia resources are classified based on at least one target feature description information of the multimedia resources, and the classification result of the multimedia resources is obtained. By adopting the invention,the accuracy of multimedia resource classification can be improved.

The invention relates to a message display method and device, a terminal, a server, a system and a storage medium. The message display method comprises the steps: receiving a session message for a target user account, and determining a service type of the session message; determining an aggregation service type corresponding to the service type of the session message; updating the session messageto a target aggregation session corresponding to the aggregation service type; and displaying the updated target aggregated session to the target user account. According to the scheme of the invention, the session message is aggregated according to the service type to which the session message belongs, so the session message can be presented in the aggregated session corresponding to the service type, a user can quickly check the session message and search for the service session meeting the service requirement of the user, and therefore, session search time can be reduced, and the man-machineinteraction efficiency is improved.

The invention provides a search method and device, electronic equipment and a storage medium, and the method comprises the steps: receiving a ubiquitous query search request; wherein the ubiquitous query search request carries ubiquitous query words; obtaining at least one multimedia content card corresponding to the generic query word; wherein each multimedia content card corresponds to one extension tag corresponding to the extensive query search request, each multimedia content card comprises information of a plurality of multimedia content sets, and each multimedia content set correspondsto one extension sub-tag corresponding to the extension tag; and displaying the information of the plurality of multimedia content sets included in each multimedia content card in the at least one multimedia content card. By the adoption of the scheme, the search results related to the generic query words can be displayed in a clustering mode from various dimensions, a user can directly find content related to the search intention of the user, search efficiency is improved, and a search path is shortened.

Methods of processing video are presented to generate signatures for motion segmented regions over two or more frames. Two frames are differenced using an adaptive threshold to generate a two-frame difference image. The adaptive threshold is based on a motion histogram analysis which may vary according to motion history data. Also, a count of pixels is determined in image regions of the motion adapted two-frame difference image which identifies when the count is not within a threshold range to modify the motion adaptive threshold. A motion history image is created from the two-frame difference image. The motion history image is segmented to generate one or more motion segmented regions and a descriptor and a signature are generated for a selected motion segmented region.

A method for classifying a multimedia content is provided. The method includes processing one or more multimedia content to obtain a set of extracted features, performing a topic modeling on the set of extracted features to obtain a set of topic models, and a set of topic keywords. Each of the topic models includes one or more explicit content topics associated with the one or more multimedia content. The method further includes identifying an explicit content topic from the topics based on the set of topic keywords, and a set of predetermined words, processing a multimedia content to obtain at least one feature, and metadata associated with the multimedia content, deriving a topic distribution based on the at least one feature and the topic models, and classifying the multimedia content as (i) an explicit multimedia content, or (ii) a non-explicit multimedia content based on the explicit content topic, and the topic distribution.

Methods of processing video are presented to generate signatures for motion segmented regions over two or more frames. Two frames are differenced using an adaptive threshold to generate a two-frame difference image. The adaptive threshold is based on a motion histogram analysis which may vary according to motion history data. Also, a count of pixels is determined in image regions of the motion adapted two-frame difference image which identifies when the count is not within a threshold range to modify the motion adaptive threshold. A motion history image is created from the two-frame difference image. The motion history image is segmented to generate one or more motion segmented regions and a descriptor and a signature are generated for a selected motion segmented region.

The invention discloses a multimedia data auditing method, device and equipment and a storage medium. The method comprises the following steps of obtaining multimedia data to be audited; extracting content feature information from the multimedia data according to the data type of the multimedia data; inputting the content feature information into a content classification model corresponding to thecontent feature information to obtain a category probability of a content category to which the multimedia data belongs; and determining the content category to which the multimedia data belongs according to the category probability. According to the embodiment of the invention, the checking efficiency and the accuracy of the multimedia data are improved.