290 results about "Paired Data" patented technology

The invention relates to the technical field of information retrieval and generation, in particular to a retrieval enhancement method based on multi-modal data fusion and modal perception. According to the method, firstly, a dual-channel architecture is adopted to perform feature extraction and coding on a text and an image respectively, and mutually independent embedded representation spaces are constructed, so that high-quality collaboration and matching of cross-modal representation are realized; and a pseudo-pairing generation mechanism is introduced to effectively mine and reconstruct the existing non-paired data in the knowledge base. And designing a query modal perception and dynamic weighting mechanism for accurately controlling the fusion proportion of the image-text bimodal information in the retrieval stage so as to match the modal demand difference of different query contents. And further executing aggregation retrieval and reordering of the cross-modal information by using dynamic weighted fusion retrieval to generate a candidate set of multi-modal responses. According to the method, accurate matching and dynamic weight adjustment of the image-text content are realized, and the accuracy and expression integrity of the generated content are improved.

The invention provides a space projection error analysis-combined 4D millimeter wave radar and visual camera automatic calibration method. The method comprises the steps of constructing a sensor internal reference model; designing a composite multifunctional calibration board and a three-dimensional adjustable holder, performing rigid geometrical relationship modeling, and automatically associating physics-features-coordinates; collecting multi-frame synchronization frame pair data, and performing time sequence fine grit alignment and interpolation compensation to obtain a space-time alignment frame pair sequence; extracting cross-modal spatial constraints of the features and automatically screening and matching point pairs to obtain a high-confidence point pair set; performing point-point constraint and point-point error, point-plane error and pixel reprojection error minimization to obtain a standardized error term; and constructing an overall objective function, and performing spatial constraint and time deviation estimation to obtain corrected external parameters and synchronization deviation. According to the invention, through a full-automatic standardized process, the probability of artificial participation and error occurrence is greatly reduced; and multi-modal and multi-scene high-precision calibration is supported, and the flexibility and robustness are greatly improved.

The invention relates to a knowledge-intensive visual question and answer automatic data generation method and device, and the method comprises the steps: constructing an original visual data set containing the professional knowledge of a target domain according to a static image, a video stream and multimedia content; extracting a representative frame sequence, converting the audio information into text information, and extracting character information in the static image to construct a structured visual instance database; according to the prompt text meeting the preset professional depth condition, establishing a three-level prompt system containing domain knowledge, an evaluation standard and a generation specification; generating a corresponding visual question and answer pair data set according to the dynamic cooperation of the main agent and the domain expert agent; generating a multi-agent quality evaluation system according to the quality evaluation result; and designing a difficulty grading mechanism according to the negative example sample. According to the method, the professionality, the accuracy and the diversity of the visual question and answer data are remarkably improved, and reliable data support is provided for training and evaluation of a multi-modal large model.

The invention discloses a cross-view image geographic positioning method based on dynamic threshold pseudo tag self-training learning, and the method specifically comprises the following steps: introducing a difficult sample feature mining method, dynamically adjusting the loss weight of a sample according to the change of similarity, and building a dynamic difficult sample triple loss model; the method comprises the following steps: dynamically adjusting a confidence threshold value of a sample by adopting an index moving average weighting method, iteratively training and screening an unlabeled sample, namely a pseudo label, establishing a pseudo label self-training mechanism of a dynamic threshold value, mining and utilizing non-paired data, and solving the problem of high manual labeling cost; a reference image most similar to a query image is found through image retrieval, and the offset of a query position is predicted. Experiments on CVUSA and CVACT data sets show that as the distance threshold increases, the accuracy of the cross-view image geographic positioning method based on dynamic threshold pseudo tag self-training learning presents a stable rising trend, and the cross-view image geographic positioning method based on dynamic threshold pseudo tag self-training learning is superior to other methods under the same threshold condition.

The invention discloses a protein generation model optimization method based on deep learning, and relates to the technical field of protein generation. The optimization method comprises the following steps: generating a candidate protein electron density distribution diagram based on the structural characteristics of a target spot by adopting a pre-trained diffusion model; converting the candidate protein electron density distribution diagram into a corresponding amino acid sequence; determining a functional index value corresponding to the amino acid sequence, and constructing a feedback data set; setting a reward threshold value, and marking the feedback data set as a positive sample data set and a negative sample data set according to the reward threshold value; constructing a utility function taking the reward threshold as a reference point, and respectively calculating utility values of the positive sample data set and the negative sample data set; and performing iterative optimization on the diffusion model according to the utility value. By adopting the technology provided by the invention, the dependence on preference paired data of large-scale and high-quality protein sequences can be avoided, and the performance of the protein generation model can be effectively and continuously improved.

The invention relates to the technical field of computer graphics and computer vision, and particularly provides a monocular relief depth estimation model construction method based on a progressive fine tuning strategy. The method comprises the following steps: constructing a pseudo real relief image data set by using a text-based image generation model; generating a depth pseudo label according to the pseudo real relief image data set; obtaining an initial relief depth estimation model based on the pseudo real relief image data set and the depth pseudo label thereof; collecting a multi-view real relief image, and generating a real relief image and depth label pairing data set in combination with three-dimensional reconstruction and geometric detail optimization; according to the method, the initial relief depth estimation model is subjected to lightweight fine adjustment by adopting low-rank adaptation LoRA, a final monocular relief depth estimation model is obtained, high-quality training data is effectively constructed in the absence of real labels, and the model generalization ability and the prediction precision are considered at the same time.

The invention discloses a laser design system, a laser design method, electronic equipment and a medium, and belongs to the technical field of integrated optics. The laser design system comprises a first model, input parameters of which comprise set grating parameters of a grating structure of a first to-be-designed laser, and which is used for predicting and outputting a predicted output characteristic curve of the first to-be-designed laser based on the set grating parameters; the input parameters of the second model comprise a set output characteristic curve of a second laser to be designed, and the second model is used for predicting and outputting predicted grating parameters of a grating structure of the second laser to be designed based on the set output characteristic curve; the first model and the second model share the same training data set, the input training data of the first model corresponds to the output training data of the second model, and the output training data of the first model corresponds to the input training data of the second model; the training data set includes pairing data between grating parameters of a grating structure of the sample laser and an output characteristic curve of the sample laser.

The invention discloses a sparse view angle 3D-DSA reconstruction method based on a three-dimensional Poisson generative model. The method comprises the following steps: obtaining pairing data of a sparse view angle 2D-DSA projection drawing and a 3D-DSA reconstruction image; extracting features of a projection image by using a projection domain encoder, then converting features of two-dimensional projection into a three-dimensional image domain according to a geometrical relationship of the cone beam CT, and then obtaining a prior image by using an image decoder; the method comprises the following steps: constructing a three-dimensional Poisson generation model, adding noise to a three-dimensional image patch in a training stage to obtain a disturbance image, outputting a network reconstruction image by taking a prior image as a condition and the three-dimensional image patch before noise addition as a target image, and calculating loss of the output image and the target image to update network parameters; the mean square error loss and the mean square error loss of the maximum intensity projection images of the three orthogonal planes are used during loss calculation; in the sampling stage, random noise is used as input, a prior image is used as a condition, noise of a noise image is continuously denoised within a limited step length, and finally a reconstructed 3D-DSA image is obtained.

The invention discloses a security constraint type inverse kinematics solving method and device and a storage medium. The method comprises the following steps: firstly, forming a joint angle-tail end pose pairing data set by students through forward movement of a robot; training an improved RBF (Radial Basis Function) network by using the data set to establish initial mapping from a tail end pose to a joint angle to obtain an initial approximation model; taking the prediction output of the model as an initial population of a PSO algorithm, and updating the speed and position of particles in the initial population through iteration; designing a comprehensive cost function, introducing a joint angle constraint and an obstacle collision constraint, and evaluating the advantages and disadvantages of each particle; and when the PSO algorithm reaches the maximum number of iterations or the comprehensive cost function value is lower than a set threshold value, outputting a global optimal particle position as a final inverse kinematics solution. On the premise of ensuring the solving precision, the joint limit constraint and the obstacle avoidance constraint are explicitly processed, so that an accurate, safe and feasible inverse solution is output, and the reliability and the safety of robot operation are remarkably improved.

The invention discloses a medical instruction data generation and use method based on a multi-modal knowledge graph. Firstly, medical images and text information in a public data set are collected, and high-quality image-text pair data are screened out from the medical images and the text information; secondly, performing entity recognition on the medical text and retrieving related triple knowledge from the multi-modal knowledge graph to form knowledge enhanced medical image-text data; secondly, inviting a medical expert to label real scene question and answer pairs based on the knowledge-enhanced medical image-text data, and taking the real scene question and answer pairs as a seed instruction data set to guide a large-scale visual language model to generate diversified instruction question and answer data; and finally, using the medical image understanding and question-answering ability of the supervised fine-tuning training model, and performing reinforcement learning supervised by a knowledge graph triple path. According to the method, the problem of scarcity of medical instruction data can be effectively relieved, so that the trained model has higher accuracy and interpretability in a medical question and answer task.

The invention discloses a face privacy protection-oriented generated image detection method and a face privacy protection-oriented generated image detection system. The method comprises the following steps of: firstly, preparing face and text pairing data, and finely adjusting a diffusion model; secondly, on the basis of the diffusion model after fine tuning, potential vectors are extracted and clustered, and text prompts and center vectors obtained through clustering form a dictionary; and finally, based on the obtained dictionary, obtaining a pseudo image and a label through the fine-tuned diffusion model, and outputting a detection result through a classifier. According to the method, potential spatial clustering and conditional diffusion generation are combined, privacy protection and data diversity are taken into consideration, and the security and generalization ability of forged face image detection are remarkably improved.

A method, an apparatus, and a non-transitory computer-readable storage medium for video compression using a generative adversarial network (GAN) are provided. The method includes obtaining, by a generator of the GAN, a reconstructed target frame based on a reference frame and a raw target frame to be reconstructed; concatenating, by a transformer-based discriminator of the GAN, the reference frame, the raw target frame and the reconstructed target frame to obtain a paired data; determining, by the transformer-based discriminator of the GAN, whether the paired data is real or fake to guide reconstruction of the raw target frame; and determining a generator loss and a transformer-based discriminator loss, and performing gradient back propagation and updating network parameters of the GAN based on the generator loss and the transformer-based discriminator loss.

The invention discloses a two-arm smart teleoperation method and system based on motion redirection and self-adaption, and relates to the technical field of robots, and the method comprises the steps: determining the posture of a human hand and the posture of a human arm according to a human body image; the robot is redirected according to the human hand posture and the human arm posture through the space attention gating map convolutional network, and the joint angle of a mechanical hand of the robot and the joint angle of a mechanical arm of the robot are correspondingly obtained; the mechanical arm is driven according to the joint angle of the mechanical arm so that the mechanical arm can grab the target object; and target force needed for grabbing the target object is determined, and the mechanical arm is driven to grab the target object according to the target force and the joint angle of the mechanical arm. According to the method, topology modeling and cross-platform action redirection between the human body and the robot are achieved through the space attention gating map convolutional network, pairing data dependence is eliminated, and therefore the high precision of action reproduction and the safety and flexibility of the grabbing process are both considered in the complex dynamic environment.

The invention belongs to the technical field of multimedia analysis, and discloses a multi-modal media tampering detection method, system and device based on multi-view comparative learning and a medium, and the method comprises the steps: obtaining a training data set which comprises a training image-training text pair and a corresponding tampering category label; a cross encoder is introduced on the basis of a vision-language model, a plurality of multi-layer sensor head structures are arranged, and three kinds of comparative learning of noise enhancement, prototype-based and multi-label tampering classification are designed to obtain an initial multi-view comparative learning framework; training the initial multi-view comparative learning framework based on the training data set to obtain a trained multi-view comparative learning framework; and based on the trained multi-view contrast learning framework, executing a tampering detection task of the to-be-detected image-text to the data. According to the technical scheme, the accuracy and robustness of multi-label classification can be improved.

The invention relates to a foreign exchange contract valuation method and device, computer equipment and a readable storage medium, and the method comprises the steps: determining a target foreign exchange transaction parameter carried by a real-time valuation request in response to the real-time valuation request generated for a foreign exchange contract; determining target hash key value pair data matched with the target foreign exchange transaction parameter from the hash key value pair data of the multi-level index data structure in the out-of-heap memory; each hash key value pair data has a corresponding data storage area; reading a plurality of foreign exchange contract data from a target data storage area corresponding to the target hash key value pair data; and performing batch calculation on the current foreign exchange data and the multiple pieces of foreign exchange contract data in a parallel mode, and determining an estimation result corresponding to each piece of foreign exchange contract data. Rapid positioning and efficient retrieval of foreign exchange contract data are realized; the problems of influence of frequent garbage collection on performance and memory leakage are avoided; and the real-time performance and the calculation efficiency of foreign exchange contract estimation are effectively improved.

The invention provides a multi-data combined construction site three-dimensional model construction method and system, and relates to the technical field of data processing, and the method comprises the steps: carrying out the calling of a preset resolution satellite image, and obtaining a full-region stereo image pair data set; after radiation correction is carried out, a region surface model and a region elevation model are generated; grid-level modeling defect detection is carried out, and distributed blind area blocks are positioned; performing obstacle avoidance path fitting to generate an obstacle avoidance optimized flight path; the unmanned aerial vehicle is driven to execute multi-angle surrounding scanning, and distributed blind area dense point clouds are directionally collected; constructing a distributed blind area model; and carrying out microscopic superposition, and outputting a three-dimensional model of the construction site. The technical problems that a construction site modeling method in the prior art is limited in the aspects of coverage, modeling efficiency and data real-time performance, depends on a single data source, cannot provide a comprehensive and accurate three-dimensional model of a construction site, and affects the construction progress and decision are solved.

The invention provides a data desensitization method which can be applied to the technical field of data security. The method comprises the following steps: receiving a data stream from a real-time data source; performing data format identification on continuously arriving data in the data stream to obtain a data format; analyzing the data by using an analyzer corresponding to the data format to obtain key-value pair data, the key-value pair data comprising respective key-value pairs of at least one field, and the key-value pairs comprising keys representing field identifiers of the fields and values representing the data of the fields; for any field in the at least one field, determining a desensitization algorithm corresponding to the field under the condition of determining that the field meets a sensitive data matching condition according to at least one of the key or the value and the sensitive data matching condition indicated by each of the plurality of desensitization rules; and desensitizing the value corresponding to the field by using an execution component corresponding to the desensitization algorithm.

The invention discloses a tactile image domain migration method and device based on a multi-scale generative adversarial network, and the method comprises the steps: constructing the multi-scale generative adversarial network, and achieving the domain migration between a simulation tactile image and a real tactile image. The generator takes U-Net as a trunk, introduces a multi-scale stacking module, a multi-stage attention gate mechanism and a channel-space attention module, and improves the reconstruction capability of simulation image textures, illumination and contact areas. And the discriminator adopts a multi-scale discrimination structure to realize the discrimination of image authenticity and detail consistency. Through joint training of joint adversarial loss, loop consistency loss, contact area consistency loss, illumination balance loss and frequency domain loss functions, it is ensured that an output image is consistent with a real image in visual and semantic levels. The method can be widely applied to a robot vision-touch fusion perception task, and the migration performance and robustness of the perception model are remarkably improved under the condition of non-paired data.

An authentication system (AS) includes a radio frequency (RF) tag reader and operates by: assigning pairing data to be associated with an RF tag and a corresponding object to be authenticated wherein the pairing data includes a tag identifier, a private key, a public key, a first random number and a first AS hash based on the first random number; pairing the authentication system to the RF tag associated with the object; receiving a request to authenticate the object; when first tag hash matches first tag hash check data, generating first read event data that indicates authentication of the object; and facilitate updating, via a blockchain system and based on the first read event data, a non-fungible token (NFT) containing a registry of read events.

The invention discloses a visual touch fusion object material identification method and device based on channel switching, and the method comprises the steps: obtaining a paired data sample which comprises an object and is composed of a visual image and a touch image, carrying out the preprocessing of the paired data sample, and constructing a training data set; constructing a visual touch fusion object material identification network, wherein the visual touch fusion object material identification network comprises a visual feature extraction sub-network, a touch feature extraction sub-network, a channel switching gating fusion module, a space attention module, a feature pyramid fusion module and a classification module; training the visual touch fusion object material identification network by using the training data set; and deploying the trained visual touch fusion object material identification network to a device end, and performing object material identification by using the trained visual touch fusion object material identification network to obtain an identification result. According to the invention, on the basis of the lightweight design of the model, the substance identification accuracy is obviously improved.

The invention discloses an equipment fault and maintenance knowledge intelligent association method based on a large model, and relates to the technical field of equipment intelligent operation and maintenance and industrial artificial intelligence, and the method comprises the steps: carrying out the structural analysis of historical fault work order data and maintenance record data of target equipment, and constructing a fault-maintenance knowledge pair data set; performing semantic coding on the fault description text and the maintenance scheme text by using a pre-trained large model, generating a corresponding semantic vector, and constructing a fault-maintenance semantic association map; and performing semantic coding on the real-time fault description, screening candidate fault nodes based on semantic similarity, calculating the semantic matching degree of the maintenance scheme in the semantic association map, and outputting a target maintenance scheme sequence after matching degree sorting. According to the invention, the intelligent level of equipment fault maintenance and the reuse efficiency of historical maintenance knowledge are improved.

The invention relates to the technical field of calculation models, and particularly discloses a petroleum knowledge question-answering method and system based on a large language model.The method comprises the steps of data availability preprocessing, question-answering pair application quality judgment and model output effect judgment, and a petroleum knowledge question-answering platform firstly collects petroleum field original data of all data sources; an available data set is obtained after availability preprocessing and verification of the data processing component; constructing a question and answer pair set according to the data set in a predefined mode, and screening out a question and answer pair data set through calibration and comparison; the platform sends a fine tuning instruction, the petroleum question and answer component finely tunes the model according to a predefined mode, evaluates question and answer output effect indexes of the petroleum question and answer component, verifies the question and answer output effect indexes with predefined indexes, judges whether fine tuning is repeated or not, and finally realizes the petroleum knowledge question and answer based on the large language model by repeating the steps.

The invention provides a data generation model training method and device, electronic equipment and a storage medium, and belongs to the technical field of artificial intelligence, and the method comprises the steps: constructing a preference data pair which comprises an initial question and answer pair and an optimized question and answer pair; inputting the preference data pair into a to-be-trained large language model, and obtaining an enhanced question and answer pair output by the to-be-trained large language model; updating the preference data pair to obtain an updated preference data pair, including: setting the enhanced question and answer pair as an optimized question and answer pair in the updated preference data pair; and iteratively executing the steps of constructing the preference data pair and updating the preference data pair until a preset termination condition is met, and obtaining a trained large language model. According to the method, the question and answer pair data is optimized by utilizing the ability of the large language model, and the large language model is trained by using the preference data pairs subjected to iterative updating, so that high-quality question and answer pair data can be generated efficiently at low cost.

The application discloses a multi-modal heterogeneous-based unmanned aerial vehicle image super-resolution reconstruction method and system and a medium, relates to the technical field of image processing, and collects multi-channel images of a first resolution and reference panchromatic images of a second resolution, generates a paired data set after preprocessing, extracts multi-channel feature maps and panchromatic feature maps from the paired data set based on a MultiHetSRNet model, generates a multi-modal heterogeneous graph by modeling nodes of a graph structure, and generates a relationship mode matrix describing spatial relationships and cross-modal relationships between nodes; target multi-channel images are reconstructed by aggregating local features and global features according to the relationship mode matrix; the scheme is improved in the method on the basis of the prior art, multi-channel images of an unmanned aerial vehicle and remote sensing reference panchromatic images are fused, complex spatial relationships are modeled by using a multi-modal heterogeneous graph, the robustness of a model is improved in combination with multiple degradation modes, and high-quality super-resolution reconstruction is realized.

The application discloses a three-dimensional medical image visual language model pre-training method of reconstruction mixed strategy, belongs to the technical field of medical image computation, and comprises the following steps: constructing a medical image text pair data set, a language text mask reconstruction strategy, a visual image mask reconstruction strategy, a semantic perception fusion strategy and multi-task joint learning; the large language model is fine-tuned, the fine-tuned large language model is used to extract diagnosis and attribute information in a medical report and generate efficient prompts, the large language model has strong generalization ability, and the cost of manual labeling is greatly saved; the semantic perception fusion strategy is used for combining text features obtained by a text encoder and image features obtained by an image encoder to obtain new text features, so that the text perceives diagnosis and attribute information of the image in advance, the alignment of the image and the text in an embedding space is further optimized, and the pre-training efficiency is improved.

Provided herein are computer-implemented methods of generating a disease map of a population. In some embodiments, the methods include applying a clustering algorithm to a set of weight and bias values of a trained electronic neural network to generate the disease map of the population. In some embodiments, the electronic neural network has been trained on training data that comprises representations of peptide sequence and binding value pair data sets obtained from reference subjects in the population in which a given peptide sequence and binding value pair data set comprises peptide sequence information and peptide binding values of antibodies to peptides that comprises the peptide sequence information. In some embodiments, the antibodies are from a sample obtained from a given reference subject in the population and are indicative of one or more disease states. Related systems, computer readable media, and additional methods are also provided.

An LZ77 decoding circuit, in the pair data flow direction, includes: an input / output control circuit, an input buffer, an input preprocessing circuit, an address mapping circuit, a RAM input / output control circuit, RAM, an output preprocessing circuit, and an output buffer. The RAM and the output buffer together constitute a lookup window. The RAM input / output control circuit searches for the data needed for the pair in the RAM of the lookup window. In the literal data flow direction, the LZ77 decoding circuit includes: an input buffer, an input preprocessing circuit, and an output buffer. Data in the output buffer updates the RAM. This invention effectively improves the LZ77 decoding speed through an improved lookup window structure and separate preprocessing of input and output data.

Systems, apparatuses, methods, and computer program products are disclosed for monitoring and tracking technical engagements between digital resources. An example method includes deploying a machine learning monitoring application to a computing environment comprising one or more of a computing device or a network channel and detecting an engagement metric associated with a target technology of the computing environment and a user account of the computing environment. The example method further includes generating name-value pair data representative of the engagement metric, the user account, and a timestamp token and generating a technical engagement score representative of a probability of a successful utilization of the target technology by the user account. The example method further includes initializing an actionable instruction set based on a comparison of the technical engagement score to a technical engagement threshold.

The invention discloses a text question-answer pair data set generation method and device, a medium and electronic equipment, and relates to the field of natural language processing. The method comprises the steps of performing hierarchical standardization analysis on a to-be-processed text, and determining a chapter logic hierarchical structure; extracting candidate knowledge points based on the chapter logic hierarchical structure, and reversely aggregating the semantic context by taking the candidate knowledge points as a core to generate a structured tetrad; positioning each structured tetrad in the to-be-processed text to obtain a text position corresponding to the structured tetrad; based on the similarity between the knowledge points in different structured tetrads, aggregating the structured tetrads meeting the similarity requirement and the corresponding text positions, and generating an index table corresponding to the knowledge points; and based on the index table, generating a question and answer pair data set corresponding to the to-be-processed text. Through an index table for associating knowledge points dispersed in different hierarchical units, the defect that a traditional method is limited to context and cannot reflect the whole knowledge structure is overcome.

The application discloses a target-level cross-modal image-text retrieval method and a storage medium. The method constructs image-target information paired data, integrates the position information of the target into the text title features in the feature extraction process, effectively learns the correlation between the target and the text in the image, and improves the accuracy and interpretability of the image-text retrieval. The cross-modal gap distribution is added to the image features, and the similarity of the image features and the target features is maximized by combining the gap cosine loss function, so that the modal gap problem of the image-text data is solved. The secondary training module is used for further training of the error retrieval data, so that a target-level cross-modal image-text retrieval model with complete training is obtained, and the performance of the model is further improved.