50 results about "Generative modeling" patented technology

The invention belongs to the technical field of seismic exploration, and discloses a robust full-waveform inversion method, system and device based on generative modeling, and a medium, and the method comprises the steps: obtaining seismic observation data; the method comprises the following steps: starting from random Gaussian noise, establishing a model space comprising a plurality of initial velocity models through an unconditional score model, and determining a global optimal initial velocity model of the model space by adopting a strategy search algorithm; de-noising is carried out on the global optimal initial velocity model based on back diffusion, seismological observation data are introduced as observation constraints, the velocity model is updated by calculating the mismatch gradient of forward modeling data of the current velocity model and the seismological observation data, and an implicit condition sample is obtained; performing forward diffusion processing on the implicit condition sample to obtain a speed model for a subsequent annealing time step; and when the annealing process reaches a preset condition, outputting a final speed model. According to the method, the robustness and accuracy of full-waveform inversion are improved, and the bottleneck problem of traditional full-waveform inversion is solved.

Text generation prompts may be determined based on an input document and a text generation prompt template. The text generation prompts may include text from the input document and questions related to the text. The text generation prompts may be sent to a remote text generation modeling system, which may respond with text generation prompt response messages including novel text portions generated by a text generation model. The text generation prompt response messages may be parsed to generate answers corresponding with the questions.

The invention belongs to the field of reinforcement learning of artificial intelligence and the field of natural language processing, and discloses a reinforcement learning-based large language model jailbreak test method, which comprises the following steps of: generating and modeling a jailbreak suffix into a sequence decision task, autonomously generating a candidate suffix through a test model, and querying a target model; and evaluating the returned text by the continuous composite reward. The continuous composite reward comprises three parts, namely (1) token-level rejection probability, (2) semantic unsafe probability of an external security model and (3) multi-anchor semantic alignment, so that continuous, dense and stable training feedback is realized. A test model is updated by adopting a Group Relative Policy Optimization (Group Relative Policy Optimization) method, so that the success rate of prison break is remarkably improved, and the query cost is reduced. The method has the advantages of being stable in training, high in cross-model generalization ability, excellent in performance on strict and safe alignment models and the like.

The system generates a three-dimensional model with layered depth images based on an input two-dimensional image. For training, layered depth images are derived from existing three-dimensional models. The system trains a machine learning model to predict multiple layered depth images from an input image of an object. The system compares the generated, multiple layered depth images to the derived layered depth images for the object to update the machine learning model during training. At inference time, the system receives an input image for an object. The system applies the machine learning model to the input image to output predicted layered depth images. The system generates a three-dimensional model from the predicted layered depth images.

The application discloses a bridge substructure modeling scheme generation method and device, equipment and medium, relates to the field of bridge engineering technology, and comprises the following steps: obtaining bridge superstructure design input parameters, constructing substructure design task information, generating design parameters by using a large language model reasoning, and then automatically generating a modeling script and performing structure response calculation. According to the calculation result, different modeling strategies are judged and executed, and finally a target modeling scheme meeting the requirements is intelligently generated, the design parameter generation and automatic modeling of the bridge substructure in a non-standard bridge type scene are realized, and the design efficiency and accuracy are improved.

A system is provided.SOLUTION: A system comprising: means for receiving a generation request; means for obtaining parameters related to a theme, a difficulty level, and a play style based on the generation request; generation model means for automatically generating a world or a dungeon from the obtained parameters; and means for storing and responding to data of the generated world or dungeon.SELECTED DRAWING: Figure 1

A system is provided.SOLUTION: A system comprising: means for receiving a creative idea input by a user; terminal means for transmitting the user's idea to a server; means for analyzing the received idea data; generative model means for automatically generating digital art based on the analyzed data; means for transmitting the generated digital art to the terminal; and means for displaying the transmitted digital art to the user.SELECTED DRAWING: Figure 1

The invention discloses an expert demonstration reinforcement learning method and system based on a visual language model and diffusion guidance. The method comprises the following steps: collecting a multi-modal environment image; performing joint coding on the multi-modal environment image and the received natural language instruction by utilizing a pre-trained visual language model to generate semantic embedding; fusing and splicing the semantic embedding and the mechanical arm state information, and inputting the fused and spliced semantic embedding and mechanical arm state information into a strategy network; performing supervision and fine tuning on the strategy network by utilizing an expert demonstration data set, and optimizing the strategy network by adopting an unconstrained PPO variant after the behavior cloning loss is minimized; in the action generation process of the strategy network, by introducing a diffusion guiding mechanism, action generation is modeled as a reverse denoising process, so that smooth continuous actions are generated; a hierarchical strategy architecture is adopted, potential sub-targets are output based on the current state, and specific actions are generated under diffusion guidance by taking the sub-targets as conditions. And the strategy generalization ability in a complex multi-modal task is improved.

This invention relates to the fields of digital core modeling and artificial intelligence modeling, and particularly to a method for generating microscopic images of weathered sandstone based on conditional generative adversarial networks (GANs). The method includes: acquiring and analyzing sandstone samples with different degrees of weathering to construct a multi-scale database of weathered sandstone; extracting weathering geological knowledge from the multi-scale database to construct a sandstone feature tag set; training the GAN using the sandstone feature tag set to obtain a microscopic image generation model for weathered sandstone; and then using the microscopic image generation model to generate physically consistent microscopic structural images of weathered sandstone. This solves the problems of existing models for modeling the physical and mechanical properties of weathered sandstone, which struggle to reveal the deep correlation between microscopic structure and macroscopic mechanical properties, and the difficulty of effectively integrating the advantages of weathering geological mechanisms and generative modeling in traditional methods, leading to a lack of physical consistency and reasonable weathering response in the reconstruction results.

The invention discloses a time sequence knowledge graph reasoning method based on autoregression condition diffusion generation, and belongs to the technical field of artificial intelligence and dynamic knowledge graphs. According to the method, entity structure dependency in a single time snapshot is aggregated through a relation-aware graph neural network, time-gated loop unit autoregression modeling is combined to capture cross-snapshot time sequence evolution characteristics, and dynamic embedding is output; simulating future uncertainty through a forward diffusion process under the condition of the embedding, and generating future representation aligned with real distribution through reverse denoising iteration to cover a multivariate potential evolution path; and designing an entity and relation fusion gate, adaptively balancing historical experience and generating a prediction weight, and combining a ConvTransE decoder to realize high-precision link prediction. According to the method, adaptive capacity to uncertainty is enhanced through generative modeling, modeling comprehensiveness is improved through multi-granularity feature fusion, history and generated information are balanced through a dynamic gating mechanism, prediction flexibility and robustness are remarkably improved, and the method is suitable for sequential reasoning tasks in multiple fields such as event prediction and a dynamic recommendation system.

A distributed generative modeling scheme employs a jointly-trained neural network path (118) implementing a generative model (112) at a UE (110) instead of wholly at a remote application server (106). An encoder neural network (120) at a base station (108) between the application server and the UE operates to, in effect, encode input prompt information (116) from the application server for efficient transmission to the UE. At the UE, a decoder neural network (122) is employed to, in effect, decode the encoded prompt information to recover a representation (126) of the original prompt information. The decoded prompt information is provided as an input to the generative model at the UE for processing, along with, in some instances, one or more additional inputs such as local sensor data (144), to generate output content (128) that is then processed by one or more UX components (114) associated with the UE.

A system is provided.SOLUTION: A system comprising: means for inputting contents to be communicated by a user by telephone; means for converting the inputted contents into text data; means for generating a dialogue script by receiving the text data and analyzing the intention; means for converting the dialogue script into voice data; means for communicating with the other party in real time by using the voice data; means for recording the contents of the dialogue and converting the recorded voice data into a text; and means for notifying the user's terminal of the summarized text.SELECTED DRAWING: Figure 1

Provided herein are methods and systems for generating infant action data and training an infant action recognition model. Video recordings for training an infant action recognition model are difficult to acquire, thus the methods and systems described may generate synthetic infant action data for training infant action recognition models with greater accuracy. The methods and systems described include a process of using real infant action recordings and corresponding infant action classifications to generate synthetic infant action data. Using an infant action recognition model, the synthetic infant action data is evaluated for accuracy. The synthetic infant action data is scored based on diversity. The synthetic infant action data may be filtered based on the accuracy and the diversity such that the filtered data represents a diverse distribution across infant action classifications. The filtered synthetic infant action data may be added to the training data to further refine the action recognition model.

This invention provides a method and equipment for park operation decision analysis, relating to the field of smart park operation management technology. The invention constructs a candidate solution scenario library through structured definitions of covariates and intervention variables, covering various operational intervention scenarios to be evaluated; relying on generative causal Bayesian networks to integrate causal inference and generative modeling capabilities, it achieves reliable causal prediction of counterfactual solutions without historical observation data; it outputs predicted effect values ​​and uncertainty indicators to quantify prediction credibility and decision risk; in high-uncertainty hypothetical scenarios, through proactive surveys and incremental updates, it achieves effect prediction of counterfactual and hypothetical solutions to assist decision analysis, reduce investment risk and decision costs, and improve the robustness and efficiency of operational decisions. This invention addresses the problem of unreliable prediction and difficulty in proactive decision-making for counterfactual and hypothetical solutions without historical data in park operations, achieving a closed-loop, end-to-end park operation decision analysis.

This application provides a data-iterative-driven multi-stage collaborative parametric modeling method for substations, relating to the field of substation digital modeling technology. The method includes: generating modeling tasks based on modeling task requests from the management end and sending them to primary and secondary professional modeling ends; acquiring nameplate rated parameter data entered by the primary professional modeling end and protection setting data entered by the secondary professional modeling end, and analyzing the protection setting data based on a parameter back-calculation analytical model to obtain back-calculated analytical data; performing collaborative iteration on the default reference values ​​of primary side parameters and nameplate rated parameter data based on a parameter collaborative iteration model to obtain iterative correction data, which is then sent to the modeling end to be corrected; receiving updated data from the modeling end to be corrected after completing the correction based on the iterative correction data, and obtaining collaborative modeling data based on the updated data, which is then sent to the management end. This application improves the real-time performance and accuracy of verifying the correspondence between various professional parameters during the multi-stage collaborative parametric modeling process of substations.

A generative adversarial network (GAN) (21, 24), or any other generative modeling technique, is used to learn (12) how to generate (68) an optimal robotic system given performance, operation, safety, or any other specifications. For instance, the specifications may be modeled (65) relative to anatomy to confirm satisfaction of anatomy-based or another task specific constraint. A machine-learning system, for instance neural network, is trained (12) to translate given specifications to a robotic configuration. The network may convert task-specific specifications into one or more configurations of robot modules into a robotic system. The user may enter (67) changes to performance in order for the network to estimate (62) appropriate configurations. The configurations may be converted (64) to estimated performance by another machine-learning system, for instance neural network, allowing modeling (65) of operation relative to the anatomy, such as anatomy based on medical imaging. The configuration satisfying the constraints from the modeling (65) may be assembled (69) and used.

A system is provided.SOLUTION: A system comprising: a data acquisition means for acquiring information of a story and a character from existing data; a generative model means for predicting an outcome of an incomplete story based on the acquired information; a formatting means for formatting the generated outcome into a text form or a visual form; and a data transmission means for providing the outcome to a user's terminal.SELECTED DRAWING: Figure 1

The invention relates to the technical field of natural language processing, in particular to an AI model-based implementation method for generating a BIM model of a transformer substation by using a language, which comprises the following steps of: generating a construction sequence by disassembling a natural language, extracting a port number and bearing strength, identifying a component type and generating a semantic parameter set; the method comprises the following steps: analyzing a statement dependency relationship to identify construction actions and parameters, determining structural strength, setting component priorities to generate a modeling scheduling list, comparing coordinates and labels to correct conflicts to generate a dependency chain sequence, generating geometries according to weights, and mapping a topological relationship to generate a BIM model, and comprises the following steps: analyzing and identifying construction actions and parameters through the statement dependency relationship, and extracting the number of ports and bearing strength for component identification; the component priority is set according to the connection features, a scheduling list is generated, conflict positioning guarantees space coordination, the modeling precision and consistency are improved through component generation binding topological relations, and efficient conversion from language to model is achieved.

Disclosed herein are methods, compositions, and devices for use in diagnosis and treatment of cancer. The methods include a generative probabilistic accounting for the characteristics of methylation data, which includes random silencing and in possesses sparsity as a result. Here, the technique finds application in subtyping, determining disease transition and formation, among other oncology applications.

A physics-based system performs generative modeling for a given dataset by translating the diffusion process in a diffusion model into a physical process. An electrical circuit provides an exemplary implementation, where each unit cell (an electrical circuit within a network of electrical circuits) consists of a resistor, a stochastic noise source (such as a thermal or shot noise source), a programmable voltage source, and a capacitor (whose charge encodes a state variable). These unit cells can be capacitively coupled with connectivity that matches the problem geometry. A score network provides the prediction of score values ​​and can be implemented on digital, analog, or hybrid digital-analog devices. The detailed structure of the analog score network is provided in the form of a physical system that evolves over time simultaneously with the diffusion process. The analog score network can continuously provide score values ​​to the backward diffusion process without delay and efficiently evaluate the loss function when connected to the forward diffusion process.

This invention discloses a plant growth simulation method based on the fusion of large-scale model generation and 3D reconstruction, belonging to the field of plant 3D visualization modeling technology. The method includes: receiving plant growth environment parameters and constructing semantic constraints for growth stages; inputting the environment parameters and semantic constraints into a multimodal large-scale model to generate a continuous, structurally consistent multi-view plant image sequence; constructing a dual path of real plant data reconstruction and generative plant reconstruction to uniformly generate plant point cloud data; inputting the point cloud data into a plant 3D modeling module, constructing a spatial representation based on a 3D Gaussian distribution, and iteratively optimizing it by combining depth-encoded geometric constraints and diffusion model appearance correction to generate a 3D plant dynamic growth visualization result driven by environmental parameters. This invention integrates the advantages of generative modeling and realistic reconstruction, achieving low-cost, high-realism plant dynamic growth simulation.

Systems and methods for training a generative neural radiance field model can include geometric regularization. Geometric regularization can involve the utilization of reference geometry data and / or an output of a surface prediction model. The geometry regularization can train the generative neural radiance field model to mitigate artifact generation by limiting a distribution considered for color value prediction and density value prediction to a range associated with a realistic geometry range.

The application discloses a BIM modeling method and device based on a multi-modal modeling code generation model, electronic equipment and a storage medium, and belongs to the field of digital modeling. The method comprises the following steps: acquiring a modeling requirement text and an image to be modeled; inputting the modeling requirement text and the image to be modeled into a multi-modal modeling code generation model, so that the multi-modal modeling code generates a BIM modeling code according to the modeling requirement text and the image to be modeled; correcting the BIM modeling code according to a preset regular expression to generate a corrected BIM modeling code; and calling a BIM modeling software, so that the BIM modeling software generates a BIM three-dimensional model according to the corrected BIM modeling code. Therefore, by implementing the application, the problem that the code deviates from the requirement due to the single modal input in the BIM modeling of the prior art can be solved.

Systems and methods described relate to the synthesis of content using generative models. In at least one embodiment, a score-based generative model can use a stochastic differential equation with critically-damped Langevin diffusion to learn to synthesize content. During a forward diffusion process, noise can be introduced into a set of auxiliary (e.g., “velocity”) values for an input image to learn a score function. This score function can be used with the stochastic differential equation during a reverse diffusion denoising process to remove noise from the image to generate a reconstructed version of the input image. A score matching objective for the critically-damped Langevin diffusion process can require only the conditional distribution learned from the velocity data. A stochastic differential equation based integrator can then allow for efficient sampling from these critically-damped Langevin diffusion models.

The application provides a kind of service recommendation model construction, service recommendation method and device, comprising: based on effective service features and derivative combination features, determine the data sample to be screened;And screen out positive samples and negative samples, generate modeling data;Utilize the initial service recommendation model of target service to predict modeling data, update the weight of negative sample according to the prediction result;Based on the updated weight of the negative sample, obtain the test data set from the modeling data, based on the test data set and the initial service recommendation model, construct the service recommendation model.The finally constituted construction service recommendation model fuses SVM model, random forest model, three-layer neural network model, XGBOOST model, and improves the generalization performance, and the constituted construction service recommendation model is used to predict the effective service features and derivative combination features of user, which can improve the success rate of service recommendation.

A text generation prompt may be determined based on an input message from a client machine and a designated text generation prompt template. A text generation prompt message including the designated text generation prompt may be sent to a remote text generation modeling system via the communication interface. A text generation prompt response message may be received from the remote text generation modeling system. The text generation prompt response message may include novel text generated by a text generation model implemented at the remote text generation modeling system. The text generation prompt response message may be parsed to generate a response text based on the novel text.

A distributed generative modeling scheme employs a jointly trained neural network path (118) that implements a generative model (112) at a UE (110) rather than entirely at a remote application server (106). An encoder neural network (120) at a base station (108) between the application server and the UE operates to effectively encode input cue information (116) from the application server for efficient transmission to the UE. At the UE, a decoder neural network (122) is employed to effectively decode the encoded cue information to restore a representation of the original cue information (126). The decoded cue information is provided as input to a generative model at the UE for processing in some cases along with one or more additional inputs, such as local sensor data (144), to generate output content (128), which is then processed by one or more UX components (114) associated with the UE.

This invention discloses a human-vehicle collaborative trajectory generation and modeling method combining multi-stage attention guidance, belonging to the field of autonomous driving technology. Trajectory prediction is achieved through a hierarchical architecture of encoding-attention-decoding. An agent state vector is constructed, and features are extracted using an embedding layer. LSTM and SLSTM encoders are used to extract temporal motion and interaction features, respectively. A multi-head attention mechanism is used to model the complex interaction relationships between agents. A decoder generates a multimodal trajectory distribution, and generative adversarial networks and diversity loss are combined to optimize trajectory generation. This invention achieves collaborative prediction of vehicle and pedestrian trajectories by combining multi-head self-attention and spatiotemporal graph attention, effectively solving the problems of insufficient interaction modeling, simplistic predicted trajectories, and fragmented human-vehicle behavior patterns in traditional methods. This significantly improves the accuracy and social plausibility of trajectory prediction in complex scenarios.

A system is provided.SOLUTION: A system comprising: means for inputting echo images; means for pre-processing the input echo images into a standard format; generative model means for generating a three dimensional model of a fetus from the pre-processed echo images; means for inputting genetic information; means for analyzing the input genetic information to predict a future appearance; and means for displaying the generated three dimensional model and the predicted future appearance.SELECTED DRAWING: Figure 1

The invention belongs to the technical field of nuclear power simulation, and particularly relates to a thermal hydraulic system simulation model generation method based on AIGC. The method utilizes a large language model to construct an intelligent agent, realizes the requirement of automatically generating a modeling code according to a user instruction, and comprises the following steps: step 1, generating a data set; step 2, fine tuning of the language model; step 3, agent construction: two agents are constructed, a task decomposition agent is responsible for decomposing a modeling task, and a modeling agent is responsible for generating a specific modeling code; and 4, performing model deployment integration. The method has the beneficial effects that on the basis of the method, the reasoning ability of a large language model can be utilized, and automatic construction from simulation system design requirements to corresponding modeling codes is achieved. Meanwhile, the generation speed of the nuclear power system simulation model can be increased, and the accuracy of conversion from design information to the simulation model can be improved.