78results about How to "Increase richness" patented technology

The invention discloses a virtual simulation data intelligent generation and enhancement method based on deep learning. The method comprises the following steps: step 1, collecting and converging multi-source heterogeneous simulation data; 2, simulation data preprocessing and feature extraction; step 3, constructing and configuring a depth generation model; step 4, iterative training and optimization of the depth generation model; step 5, intelligently generating and synthesizing new simulation data; step 6, generating data quality verification and enhancement feedback; step 7, carrying out controllability adjustment and directional enhancement on the generated data; and 8, model online learning and data generation closed-loop optimization are carried out. The method can automatically learn the internal law and distribution of the simulation data, efficiently generate highly-vivid and diversified new simulation data meeting specific requirements, and deeply enhance the existing limited data, thereby significantly reducing the simulation data acquisition cost, improving the data quality and richness, and improving the simulation data acquisition efficiency. And powerful data support is provided for development and test of a virtual simulation system and simulation-based AI model training.

This specification provides a method, device, medium, and program product for training a generative model and retrieving goods. The method includes: acquiring first sample data including first sample goods information; generating multiple first sample query messages from an initial generative model based on the first sample goods information; determining a first positive sample and a first negative sample from the multiple first sample query messages; wherein the correlation between the first positive sample and the first sample goods information is higher than the correlation between the first negative sample and the first sample goods information, and the number of historical queries corresponding to the first positive sample is greater than the number of historical queries corresponding to the first negative sample; determining a first probability that the initial generative model outputs the first positive sample based on the first sample goods information and a second probability that the initial generative model outputs the first negative sample based on the first sample goods information, and training the initial generative model with the goal of increasing the first probability and decreasing the second probability to obtain a target generative model.

The application discloses a social message display method and device, equipment and storage medium, and belongs to the field of human-computer interaction. The method is executed by a first client, and the method comprises the following steps: receiving a first social message sent by a first account in a chat session, displaying the first social message in a first view layout; and displaying the first social message in a second view layout under the condition that a first trigger condition is met; wherein at least one of the size, the layout position of the second view layout and the first view layout and the account information of the first account is different. The display mode of the view layout of the social message is changed after the trigger condition is met, the social message in the chat session can be dynamically displayed, the display mode of the social message is enriched, and the display efficiency of the social message is improved.

The application provides an edge device-oriented lightweight multi-modal early gastric cancer collaborative auxiliary diagnosis system, which comprises a lightweight multi-modal classification module, a decision-level fusion unit, a first feature extraction network and a second feature extraction network; the decision-level fusion unit is used for evidence theory fusion of first classification probability and second classification probability, and outputs a multi-modal classification result; a lightweight segmentation module is constructed based on an encoding-decoding structure, an encoder of the lightweight segmentation module comprises a plurality of stacked lightweight bottleneck modules, a feature enhancement module is arranged in a skip connection of the encoder, and the feature enhancement module is used for pixel-level lesion segmentation of a white light endoscopy image; a gating cooperation unit is connected with the lightweight multi-modal classification module and the lightweight segmentation module respectively, and is used for generating a gating signal according to the multi-modal classification result; when the multi-modal classification result is positive, the lightweight segmentation module is activated to perform a segmentation task; and when the multi-modal classification result is negative, a subsequent segmentation process is terminated.

The present disclosure provides a search result sorting method and device, computer equipment and a storage medium, wherein the method comprises: obtaining search results matched with a search term, the search results comprising first results of a first genre and at least one second result of a second genre other than the first genre; the second result comprising at least one result document; for any result document in the second result, determining a document feature corresponding to the result document according to text attribute information of the result document, dimension information of the result document in a plurality of preset characteristic dimensions, the search term and source information of the search term; determining a document association feature corresponding to the second result according to the document feature of each result document in the second result; and sorting the first results and the second results according to the document association feature of each second result and a result click rate of each first result.

The application relates to the computer technical field, in particular to a page processing method and device, electronic equipment and storage medium, adopting a standard language, converting at least one original jump component based on an original language contained in an original application page into corresponding target jump components respectively; determining page jump indication information between the original application page and each other application page based on the at least one original jump component; adopting the standard language, converting at least one original page element based on the original language into corresponding target page elements based on the element conversion mode corresponding to each original page element contained in the original application page; and sending the obtained at least one target jump component, page jump indication information and at least one target page element to the search server for associated storage. In this way, the richness of search results can be improved.

The application relates to an industrial quality prediction method and equipment based on a graph attention long short-term memory network, which comprises the following steps: acquiring a data set to be predicted in an industrial manufacturing process, performing a standardization operation on the data set to obtain a standardized data set; acquiring a low-order feature variable matrix of the standardized data set through a multilayer perception machine, and constructing an adjacency matrix; inputting the standardized data set, the adjacency matrix and the low-order feature variable matrix into a trained graph attention long short-term memory network model, splicing time feature information and space feature information output by the model through a quality estimator, and obtaining a quality prediction result; the graph attention long short-term memory network model comprises a graph attention layer and a long short-term memory layer arranged in parallel, and is used for extracting the time feature information and the space feature information, respectively. Compared with the prior art, the application has the advantages of fusing time-space features, automatically constructing an adjacency matrix and improving prediction accuracy.

The invention relates to a live broadcast content recommendation method and device, electronic equipment and a storage medium. The method comprises the following steps: displaying a content recommendation card on a live broadcast guide page; the content recommendation card comprises a recommendation tag and at least one piece of recommendation content corresponding to the recommendation tag; and in response to a trigger operation on the content recommendation card, displaying a live broadcast room, and displaying a corresponding content page in the live broadcast room based on the number of the recommended contents. According to the method, the recommendation dimension can be expanded, display of multiple recommendation contents under different recommendation tags is realized, the content currently explained by the live broadcast room is not limited to be displayed, and the richness of content recommendation is improved.

This utility model relates to the field of display stands, and discloses a surround display stand, including a central stand, a horizontal plate fixedly connected to the inner center of the central stand, a top stand on the top of the central stand, and side stands connected to the left and right sides of the central stand by limiting components. Each limiting component includes a retaining frame fixedly connected to the back of the side stands, with a limiting groove on the outer wall of the retaining frame. A limiting block is elastically connected to the inner wall of the central stand by a limiting spring, and a limiting rod is fixedly connected to the outer wall of the limiting block. The left and right sides of the central stand have retaining slots. In this utility model, the limiting components allow for flexible assembly and disassembly of the side stands and the central stand. Furthermore, the layered structure design of the central stand, side stands, and top stand allows for reasonable layout of products of different sizes and types, making full use of space for product display, greatly enhancing the richness and attractiveness of product displays, and meeting diverse display needs.

This invention relates to a method for preparing a soybean extract for the pretreatment of cigar-type reconstituted tobacco raw materials, as well as its products and applications. The method includes the following steps: roasting soybean raw materials at 150-210℃ for 5-10 minutes; then roasting at 220-250℃ for 7-12 minutes; and finally roasting at 260-300℃ for 12-20 minutes to obtain roasted soybeans; pulverizing the roasted soybeans and then hot-extracting them to obtain an extract; and then subjecting the extract to sedimentation to remove impurities. In the soybean extract obtained by this method, large molecular carbon and nitrogen compounds such as starch decompose upon heating to generate a large number of small molecular sugars and amino acids, and react to produce a large number of aroma precursors such as glycoamino condensates. These aroma precursors serve as substrates for microorganisms during tobacco fermentation, making the nutty, roasted, sweet, and bean aromas of the fermented tobacco more prominent, further enhancing the flavor and aroma of the fermented tobacco.

The invention relates to the technical field of artificial olfaction and intelligent sensing, and discloses a gas recognition method and system based on a physical-data dual-drive framework. The method comprises the following steps: acquiring a macroscopic dynamic sensing signal triggered by target gas on a sensor array containing atomic-scale catalytic active sites, and a microscopic physical descriptor for representing an atomic-scale electron interaction rule between gas molecules and the catalytic active sites; and the data branch network and the physical branch network are respectively input into the data branch network and the physical branch network for feature extraction and mapping, cross-modal fusion is carried out through a feature fusion sub-module to obtain joint feature representation constrained by a physical rule, and accordingly, a gas category identification result is output. According to the method, the microcosmic physical descriptors are fused into the model reasoning logic, so that the recognition precision of the structural similar object is effectively improved.

The application discloses a planting method for improving tobacco quality, and the fertilization mode of the planting method is set as T3, or T4, or T5. The fertilization treatment is beneficial to tobacco plant growth, tobacco carbon and nitrogen metabolism, tobacco leaf quality, and improves tobacco leaf yield.

The application discloses a software upgrade time prediction method, device, equipment and medium. The software upgrade time prediction method comprises the following steps: acquiring a plurality of historical relative time data, wherein each historical relative time data represents the relationship between the time required for software upgrade at the vehicle end and the software data size; obtaining target sequence data based on the plurality of historical relative time data and a plurality of data numbers corresponding to the plurality of historical relative time data; performing data processing based on the target sequence data to obtain a time prediction model; predicting the target relative time data required for software upgrade at the vehicle end based on the target data number by using the time prediction model; and sending the target relative time data to the vehicle end, so that the vehicle end determines the time required for software upgrade based on the target relative time data and the software data size. The time prediction model obtained through data processing can improve the accuracy of the predicted software upgrade time.

This utility model discloses a tea drying machine, belonging to the field of tea processing technology. It includes a drying cylinder with a top cover on its top surface and multiple discharge ports on its outer surface. A drive motor is fixedly installed on the inner wall of the bottom surface of the drying cylinder. This utility model incorporates a tea-turning component and a tea-holding component. This design, using a conical receiving structure, allows tea leaves to accumulate on the surface of the receiving structure after being added. Combined with the turning structure, real-time and comprehensive turning of the tea leaves during the drying process is achieved, greatly improving the uniformity and effectiveness of drying. Simultaneously, a dispensing plate near the surface of the receiving structure can also work synchronously, effectively dispensing tea leaves that are damp and adhered to the surface, further improving the drying effect and uniformity of the tea.

This invention belongs to the field of pharmaceutical technology, and relates to the use of a combination of sour cherry powder and celery seed powder, and complexes containing this composition. Specifically, the sour cherry powder and celery seed powder composition is used in the preparation of drugs for treating or regulating intestinal flora imbalance caused by hyperuricemia. The preparation method of this composition is as follows: pitting sour cherries, freezing them separately with liquid nitrogen, grinding them into powder using a pulverizer, and sieving them separately to obtain sour cherry powder and celery seed powder; mixing the obtained sour cherry powder and celery seed powder at a ratio of (0.5~3):1. This invention discusses the diversity, richness, and evenness of intestinal flora through diversity analysis, community bar chart analysis, and Venn diagram analysis. The results show that this composition can increase beneficial bacteria (Prevotella, Mulliez, Lactobacillus, etc.), reduce harmful bacteria (Alternaria), improve the structure of intestinal flora, and maintain the balance of intestinal flora.

This disclosure provides a game skill interaction method, a game skill interaction device, a computer storage medium, and an electronic device, relating to the field of game technology. The method includes: providing a first virtual character in a current game scene area displayed in a graphical user interface, the first virtual character being controlled by a terminal device; responding to a first touch operation on a skill trigger control in the graphical user interface, displaying a first target scene area of ​​the game scene in the graphical user interface, and generating a virtual clone of the first virtual character in the first target scene area; controlling the virtual clone to pick up skill items in the first target scene area, and using the skill items picked up by the virtual clone as triggerable skill items for the first virtual character. This disclosure can improve the interactivity between a player-controlled game character and other game characters regarding game skills.

The application discloses a kind of photovoltaic panel intelligent cleaning trolley stain identification and accurate cleaning method, belong to photovoltaic operation and maintenance technical field, including cleaning trolley initialization and moves to photovoltaic panel starting detection position, completes attitude calibration and detection system preheating, the present application is by fusing multispectral image acquisition, deep learning identification model and self-adapting cleaning control strategy, constructs a set of set stain identification, classification determination, path planning and accurate cleaning in one integrated intelligent cleaning method, realizes the high-precision identification of photovoltaic panel surface stain and differentiates accurate processing, significantly improves cleaning effect and energy utilization efficiency;Stain is divided into mild, moderate, severe three grades, which provides a scientific basis for subsequent differentiated cleaning;Overall identification process is fast, stable, supports real-time processing, which greatly improves the intelligent decision-making ability of cleaning system.

The application provides an image sensor and an imaging device. The image sensor comprises a microlens array, a color filter array and a photosensitive pixel array. The microlens array comprises a plurality of first lens units and at least one second lens unit. The color filter array comprises a plurality of first color filter units and at least one second color filter unit. The transmittance of the first color filter units is different from that of the second color filter units. The first lens units are arranged correspondingly to the first color filter units, and the second lens units are arranged correspondingly to the second color filter units. The photosensitive pixel array comprises a plurality of pixel units. The first color filter units cover at least one pixel unit, and the second color filter units cover at least one pixel unit. Light signals passing through the first lens units and the first color filter units form first electric signals. Light signals passing through the second lens units and the second color filter units are used to form at least second electric signals. An image signal is formed based on the first electric signals and the second electric signals.

This application presents a method for providing clear, accurate, and intelligent AI representations of complex, multi-element images. It employs an improved Faster R-CNN model as the image processing model, replaces VGG-16 with DenseNet for feature extraction, optimizes bounding box selection through dimensional clustering, and enhances model stability by predicting relative grid cell coordinates. A text generation architecture is constructed based on a hierarchical recursive network. The first layer, a global long short memory layer, determines the topic vector based on the fusion vector to increase the number of descriptive sentences. The second layer, a sentence BRNN layer, generates accurate and rich descriptive sentences. A fully convolutional localization layer effectively integrates and transfers block features, constructing an end-to-end architecture that collaboratively optimizes image processing and text generation. This method addresses the problems of high model complexity, inaccurate bounding box localization, and insufficient association of descriptive sentences in existing technologies, generating denser, more accurate, and natural descriptive language that conforms to language habits. It is suitable for various application scenarios such as visual intelligent question answering and video subtitle generation.

The invention relates to the technical field of artificial intelligence, and particularly discloses a health management scheme generation method and device based on artificial intelligence. Performing entity recognition and intention recognition on the user voice to obtain disease features and user intention; obtaining target health data associated with the disease characteristics; analyzing the disease characteristics and the target health data to obtain disease information and predict health risks; and generating a health management scheme based on the user intention, the disease information and the predicted health risk. The health management scheme is generated in combination with the user intention, the individuation degree is improved, the health parameters associated with the disease features are obtained, the data richness is improved, and then the accuracy of the health management scheme is improved. When the method is applied to health management systems such as medical care, family health monitoring and the like, a health management scheme with high individuation degree and high accuracy can be generated according to the intention of the user and the specific health data, so that the user experience is improved.

The embodiment of the application provides a data acquisition method and equipment for a humanoid robot training field, and belongs to the technical field of industrial manufacturing equipment of the humanoid robot training field. The method comprises the following steps: in response to receiving a task instruction, determining a target cleaning position on a to-be-cleaned object; based on a preset weight function, performing trajectory planning according to the target cleaning position to obtain a target trajectory; the preset weight function is associated with a first distance and an arm angle; the first distance is the distance between the target cleaning position and an end effector; performing a cleaning task on the target cleaning position according to the target trajectory; collecting and uploading target data in the task execution process; the target data comprises at least one of the following: the target cleaning position and the target trajectory. The method provided in the embodiment can improve the data acquisition efficiency and data accuracy of the humanoid robot in the training field, and can help improve the work efficiency and safety in the industrial manufacturing industry or feature work scene.

The application relates to the technical field of stem cell culture, and provides a fusion degree evaluation method and equipment for stem cell expansion culture, which comprises the following steps: calculating a microenvironment feature by using a microenvironment parameter and a stem cell fusion degree parameter; obtaining a time perception label by performing time perception based on a two-dimensional image, and obtaining a fusion degree prediction label by performing fusion degree classification; fusing a two-dimensional image and a three-dimensional image based on the time perception label and the fusion degree prediction label to obtain an enhanced stem cell image; performing time perception on the enhanced stem cell image to obtain an enhanced time perception label, and performing fusion degree evaluation to obtain an enhanced fusion degree prediction label; and performing fusion degree evaluation by using a fusion degree evaluation model based on the microenvironment feature, the time perception label, the fusion degree prediction label, the enhanced time perception label and the enhanced fusion degree prediction label, so as to obtain a fusion degree evaluation value. The method can improve the accuracy of fusion degree evaluation.

The present disclosure provides a game plot script generation method and device, a computer storage medium and an electronic device, and relates to the technical field of games. The method comprises: receiving game plot information based on a natural language description for a target game plot, and converting the game plot information based on the natural language description into structured plot requirement information; obtaining a pre-constructed plot constraint condition; inputting the structured plot requirement information and the plot constraint condition into a pre-trained large language script generation model to generate a function call sequence; and generating a target plot script file based on the function call sequence, the target plot script file being used for analysis and execution by a game engine to drive plot animation performance of the target game plot. The present disclosure can reduce the production threshold of the game plot script, thereby reducing the plot development complexity, reducing the development cost of the game plot production process, and improving the production efficiency of the game plot.

The application relates to a Transformer-based power multi-modal full-element sample fusion labeling method and system, which comprises the following steps: S1: obtaining multi-modal original data of a production operation site, and obtaining a preprocessed multi-modal data set through preprocessing; S2: performing multi-modal feature coding and Transformer representation modeling according to the preprocessed multi-modal data, and obtaining a fusion feature representation through cross-modal feature fusion and information completion; S3: acquiring a multi-modal semantic alignment label set through semantic consistency detection and a semantic mapping mechanism according to the fusion feature representation and a corresponding candidate label set; S4: constructing a label system structure tree according to the multi-modal semantic alignment label set; and S5: constructing an automatic labeling module to generate preliminary labeling based on the fusion feature representation and the label system structure tree, and directly inferring labels based on similar samples and category probabilities. The application realizes efficient automatic label generation and sample dynamic classification.

The application provides a sports event data management and intelligent broadcasting device with multi-modal input, comprising a smart control unit, an external input interface and an external output interface, wherein the smart control unit is configured with a central event control module, an automatic action execution module, a shared memory, a frame reading process and a display rendering process; the central event control module is configured to receive an event trigger signal from the external input interface; the frame reading process is configured to continuously acquire original video frames from a video input source, pre-process the original video frames and write the processed frame data into the shared memory; and the display rendering process is configured to asynchronously read the frame data from the shared memory and render the frame data to the external output interface by using a graphics rendering engine. The application provides a sports event data management and intelligent broadcasting device with multi-modal input, which is realized based on a general computer platform and by using a multi-process software architecture.

A mouse toy used in a rodent cage, relates to the technical field of barrier environment experimental animals, comprising a pipe body, sleeves are arranged on both sides of the pipe body to meet the biting habit of mice, a portable dismounting mechanism is arranged between the pipe body and the sleeve for plug-in fixing of the pipe body and the sleeve, an escape channel is arranged in the pipe body for back-and-forth shuttling movement of the mouse, an access channel is arranged in the sleeve and communicates with the escape channel, the scheme not only improves the animal welfare of rodents in the daily feeding process, but also maximizes the reduction of investment, cost reduction and efficiency increase.

This application relates to a spatial transcriptome gene expression prediction method. The method includes: acquiring stained pathological image data of a target tissue; dividing the stained pathological image data into image blocks to obtain at least two image blocks to be identified; for each image block, determining a first cross-modal image feature of the image block and a second cross-modal image feature of the adjacent image blocks; predicting the image block based on the aggregation result of the first and second cross-modal image features to determine the gene expression prediction result and cell type prediction result corresponding to the image block; and generating a spatial transcriptome gene expression prediction map of the target tissue based on the gene expression prediction result and cell type prediction result of each image block. This method can improve the prediction accuracy of gene expression profiles.

The invention relates to the technical field of data analysis, in particular to a news propagation influence quantitative evaluation method and system based on data mining, and the system comprises a data collection and preprocessing module, a dynamic index management module, an intelligent modeling and analysis module, a causal inference and explanation module and a storage and retrieval module. In order to solve the problem that a traditional evaluation model is insufficient in scene adaptability due to the fact that a fixed weight and a linear attenuation hypothesis are adopted, a dynamic attenuation model and a reinforcement learning mechanism are introduced in the scheme, and chain reactions in the information spreading process are described through modeling of a self-excitation effect in a spreading event; on the basis, an evaluation index weight distribution mechanism is dynamically optimized in combination with an adaptive algorithm, so that the model can automatically adapt to rapid outbreak of sudden news and long-acting propagation characteristics of policy news, and a nonlinear rule presented in the propagation process is accurately captured; according to the method, the scene adaptability and timeliness of the evaluation result are effectively improved.

The present disclosure provides an information processing method and program product, providing a graphical user interface through a terminal device, wherein the graphical user interface includes a game scene corresponding to a card game; the method comprises: in response to the end of the card game, determining a key card combination used by the opponent in the card game; controlling the graphical user interface to display review information of the key card combination; determining a recommended counter card for the key card combination; controlling the graphical user interface to display the recommended counter card and providing an adding option for the recommended counter card. The operation process of the player's review and countermeasure seeking is greatly simplified, the key to failure can be intuitively understood, and clear improvement suggestions can be obtained, thereby significantly improving the interactive experience. Secondly, by intelligently recommending counter cards, the player is guided to try new card combinations and tactical ideas, breaking the player's inherent card set building habits, enriching the strategy dimension and playability of the game, and effectively improving the richness of the game.

This application provides a video generation method, apparatus, electronic device, computer-readable storage medium, and computer program product. The method includes: determining text information for describing the video; identifying words describing the foreground in the text information; identifying second noise data belonging to the foreground in first noise data of the video; and generating the video based on the words describing the foreground, the first noise data, the text information, and the second noise data. This application can enrich the details of the foreground and background in the video, improving the quality of the generated video.