85 results about "Neural fields" patented technology

The invention discloses a fault diagnosis and early warning method for an intelligent thermal management system of a power battery, and relates to the technical field of new energy automobile power battery safety. Comprising the following steps executed in sequence: S1) based on real-time monitoring data of a distributed temperature sensor in a battery module, constructing a dynamic space-time neural field model through an adaptive graph neural network, and generating a thermal field fingerprint representing a thermal propagation topology mode; according to the fault diagnosis and early warning method for the intelligent thermal management system of the power battery, a dynamic space-time neural field model is constructed, a sub-DEG C weak thermal abnormal signal is accurately captured, and a micro short circuit fault is recognized in advance in combination with electric heating delay analysis, so that the problem that a fault precursor signal is submerged by a complex working condition is solved; a double-channel verification mechanism depends on double mutual verification of physical law rigid constraint and intelligent residual feature recognition, the misjudgment risk caused by cooling system fluctuation and sensor noise is eliminated, and the early warning accuracy is improved.

The invention provides a thyroid cancer auxiliary diagnosis and metastasis risk prediction method based on deep learning, and relates to the technical field of artificial intelligence auxiliary medical treatment, and the method comprises the steps: extracting ultrasonic image multi-scale features through a self-adaptive neural architecture search network, combining clinical examination data, fusing diagnosis and treatment knowledge through a neural symbol inference device, and carrying out the prediction of the metastasis risk. Generating a knowledge enhancement feature map; constructing a feature propagation field by using a dynamic neural field network, solving a dynamic evolution equation, and generating a spatial-temporal feature field representing the dynamic change of focus features; constructing a tumor diffusion kinetic model by using an implicit neural representation network and a nerve ordinary differential equation network, calculating a transition probability based on an optimal transmission algorithm, solving an optimal control equation, and outputting a metastasis risk prediction result of each organ; the thyroid cancer diagnosis accuracy and metastasis risk prediction reliability can be effectively improved, and doctors can be assisted in accurate diagnosis and treatment.

The invention belongs to the field of visual tracking, and relates to a multi-target visual tracking method based on cooperation of a dynamic neural field and a neural network, which takes a cross-modal cooperation architecture as a core and comprises a dynamic neural field module based on multi-target trajectory maintenance and shielding matching and an improved MoESDQ neural network module. Meanwhile, a collaborative decision-making mechanism is designed, when the activation peak value of the dynamic neural field is attenuated to a preset threshold value, neural network feature matching is triggered, and disappearance target reproduction correlation is achieved based on cosine similarity. The objective of the invention is to solve the visual tracking capability under the condition of scene and target motion change in a monitoring range, for example, under an intelligent traffic intersection scene. The problems of high ID switching rate, multi-target misassociation and low tracking precision under a real-time tracking background caused by scene change or frequent shielding of vehicles and pedestrians, similar target appearances, transient disappearance and reproduction of the targets and sudden illumination change are solved.

The invention discloses a three-dimensional wind field reconstruction method and system based on deep learning multi-source meteorological data. The method comprises the following steps: collecting wind field data through a Doppler laser radar and a wind profile radar; based on an MLP as a trunk architecture, introducing position coding, and constructing a neural field model; and performing zero-sample super-resolution prediction on the wind field data based on the neural field model to generate a reconstructed three-dimensional wind field. According to the method, the complex space-time correlation in the wind field can be adaptively learned, and the space-time continuity of the reconstruction result is ensured through overlapped boundary processing. The three-dimensional wind field reconstruction method and system based on the deep learning multi-source meteorological data can be widely applied to the technical field of three-dimensional wind field reconstruction.

A building-integrated artificial intelligence system forming a continuous ambient neural field is disclosed. The system includes a distributed multimodal sensor lattice, an on-premise symbolic cognition engine, and an adaptive environmental control kernel operating entirely at the building edge without reliance on external cloud services. Sensor data from optical, thermal, acoustic, airflow, pressure, structural, electrical, and chemical modalities are transformed into non-identifying occupancy vectors, behavioral glyphs, risk indicators, and environmental state descriptors. A privacy-governed policy graph determines sensor permissions, redaction thresholds, consent conditions, emergency overrides, and jurisdiction-specific compliance parameters. The neural field predicts occupancy loads, optimizes HVAC, ventilation, and lighting, detects accidents and structural anomalies, classifies emergent risks, and generates redacted event capsules for audit and emergency dispatch. A federated topology enables multiple buildings to exchange compressed symbolic templates to improve predictive accuracy without transmitting raw data. The system provides a universal, regulation-aligned AI nervous system for autonomous building operations.

The invention discloses a cultural relic restoration decision-making method and system based on computer three-dimensional modeling, and relates to the technical field of cultural relic restoration, and the method comprises the following steps: building a unified time baseline, generating a phase reference field, introducing a double-frequency comb interference structure to reconstruct a humidity refraction field, locking a zero-phase anchor point, and forming an initial constraint condition of a scanning path; under an initial constraint condition, an airflow phase gradient is obtained by adopting a Schlenn pilot frequency chromatography method, a nonlinear diffraction nucleus is calculated, and a refraction residual spectrum is output for modeling correction. According to the method, humidity disturbance is recognized through interference measurement and chromatography, real-time compensation is achieved in combination with phase sheath and humidity control, the crack and material characteristics are fused through neural field modeling, the solidification process is played back in combination with a mineralization mechanism, and the three-dimensional modeling precision and the restoration decision scientificity are improved.

The invention relates to the technical field of intelligent mines, and discloses a geological guarantee intelligent planning mining system based on deep reinforcement learning, which comprises a geological guarantee basic module, an intelligent planning engine module and a closed-loop execution system module, the geological guarantee basic module constructs a dynamic probabilistic digital twinning environment through a geological uncertainty implicit neural field, and realizes data real-time fusion updating in combination with a gradient flow; the intelligent planning engine module realizes long-period risk perception planning based on a geological perception Markov decision process and a causal world model; the closed-loop execution system module deploys a lightweight model through strategy distillation, establishes a twinborn reality bidirectional evolution closed loop, and realizes planning and execution adaptive collaboration; according to the method, the problems of static fragility of geological guarantee, short view non-strategy planning and system splitting open loop of a traditional mining system are solved, full-life-cycle dynamic cognition, deep planning and closed-loop self-evolution of a mine are achieved, and mining benefits and safety are remarkably improved.

The invention provides a Suzhou dialect speech recognition system and method based on a tone track neural field, and the system comprises a tone track neural field module which is used for modeling the tone change of Suzhou dialects into a continuous space-time neural field; the bidirectional semantic memory network module comprises a forward prediction memory bank and a backward correction memory bank; the phoneme-font coupling error corrector is used for realizing polyphone disambiguation and homonym error correction by establishing association mapping between a phoneme sequence and a font sequence; the semantic entropy calculation module is used for evaluating the uncertainty of the recognition result; and the self-adaptive fusion decision module is used for generating a final recognition text. According to the invention, through an online learning mechanism, the system can continuously accumulate experience from actual use, automatically discover a new language mode and update an identification strategy. The self-improvement capability enables the system to adapt to the dynamic change of languages, and the performance is continuously improved along with the increase of the use time. Each use of the user helps the system to become more intelligent and accurate.

The invention discloses a bionic self-crimping material and application in the nerve field, and relates to the technical field of nano materials. The preparation method comprises the following steps: firstly, carrying out ultraviolet photopolymerization on acrylic acid and acrylic acid-N-hydroxysuccinimide, adding chitosan and glutaraldehyde to form hydrogel, and introducing epigallocatechin gallate into a chitosan-acrylic acid network. The material is rapidly and autonomously curled into a tube shape after being in contact with wet nerves, glutaraldehyde is directionally and orderly diffused by adopting a three-channel micro-fluidic chip, and the fineness is improved, so that the matching degree of nerve broken ends with different thicknesses is increased; the surface active groups of the material and the natural component chitosan cooperate to realize suture-free fixation; the plant phenol antioxidant removes free radicals at nerve injury parts; the degradation period is synchronous with natural healing of nerves, and secondary operation for taking out is not needed. The electroneurographic signals using the material are transmitted more smoothly, the target muscle atrophy degree is reduced, and the limb movement function is recovered to be close to the normal level.

The invention provides a Gaussian neural field dynamic scene reconstruction system based on depth consistency constraint, and relates to the technical field of computer graphics, and the system comprises an estimation module which generates a target frame initial depth map; the calculation module reconstructs the point cloud and obtains a point cloud normal direction and a pixel normal direction; the optimization module is used for iteratively correcting the initial depth based on the two types of normal consistency to obtain an optimized depth map; the alignment module is used for determining a scale parameter through regression by taking the first target frame as a reference, and carrying out scale transformation on the depths of other frames to form a consistent depth sequence; the reconstruction module is used for constructing or training a Gaussian neural field based on the sequence and outputting a three-dimensional representation; in addition, the calculation module can contain multi-dimensional wavelets and sparse reconstruction and is used for multi-scale noise suppression and direction weighted fitting. Reference frame selection is based on frame-level quality, geometry and scale stability indexes; according to the system, the intra-frame geometric credibility and the cross-frame scale consistency are improved, ghosting and tearing are reduced, and the stability and integrity of dynamic scene reconstruction are enhanced.

The invention provides a five-axis rough machining path planning method based on a volume neural field. The five-axis rough machining path planning method comprises the following steps: S10, constructing a body region and outputting the neural field; s20, carrying out layering and path orthogonality and density constraint; s30, performing continuous and periodic smoothing processing on the cutter shaft field; s40, calculating collision safety and a minimum gap constraint; s50, the cutting volume is subjected to stabilizing treatment; s60, calculating total loss and determining a training process; and S70, tool path generation and post-processing are carried out. According to the scheme, the hierarchical field psi, the path field phi and the cutter shaft angle fields (A and C) are learned in the machining volume through the neural network, geometric orthogonality, cutter shaft smoothing / amplitude limiting, collision safety and removal stability are restrained in a unified mode through a micro-loss function, and therefore efficient, smooth and safe automatic rough machining path generation is achieved in complex structures such as a deep cavity and an impeller. The rough machining effect is good.

This application provides a Gaussian neural field dynamic scene reconstruction system based on depth consistency constraints, relating to the field of computer graphics technology. The system includes: an estimation module that generates an initial depth map of the target frame; a calculation module that reconstructs the point cloud and obtains the point cloud normal and pixel normal; an optimization module that iteratively corrects the initial depth based on the consistency of the two types of normals to obtain an optimized depth map; an alignment module that, using the first target frame as a reference, determines the scale parameters through regression and performs scale transformation on the depths of the remaining frames to form a consistent depth sequence; a reconstruction module that constructs or trains a Gaussian neural field to output a three-dimensional representation based on this sequence; in addition, the calculation module may include multi-dimensional wavelets and sparse reconstruction for multi-scale noise reduction and direction-weighted fitting of normals; the reference frame selection is based on frame-level quality, geometric, and scale stability indices. This system improves intra-frame geometric reliability and cross-frame scale consistency, reduces ghosting and tearing, and enhances the stability and integrity of dynamic scene reconstruction.

A method for performing deformable image registration of a first volumetric medical image to a second volumetric medical image may comprise estimating a time varying velocity field between the images by encoding coordinates using a time varying positional embedding and using the encoded coordinates and a Neural Field Ordinary Differential Equation (NFODE) to generate a prediction of the rate of change of the deformation field. The method may further comprise integrating the estimated velocity field to generate a deformation field and applying the generated deformation field to the first volumetric medical image to generate a registered volumetric medical image. The NFODE May comprise a non-stationary Neural ODE parameterized by an Implicit Neural Representation.

The invention provides a micro-nano structure three-dimensional reconstruction method and system based on a scanning electron microscope, and belongs to the fields of biological micro-nano surface structure analysis, ultra-precision machining and manufacturing and the like. The method comprises the following steps: acquiring an SEM image of a multi-view multi-detector of a to-be-measured micro-nano sample, performing initial geometric estimation based on a motion recovery structure algorithm, taking a neural field model as an implicit representation model of a three-dimensional shape, introducing a differentiable BSE forward model, and realizing reconstruction through a three-stage training strategy. According to the method, the neural field is used as a representation basis of three-dimensional reconstruction, so that the reconstruction capability of a complex and discontinuous microstructure is remarkably improved; a standard sample does not need to be used for parameter calibration, reconstruction can be completed only by depending on a group of SEM images of the target sample, and the operation process is simplified; the method has robustness to the shadow area in the image, effectively reduces the interference of the shadow on the surface reconstruction result through the shadow separation strategy introduced in the training process, improves the reconstruction accuracy, and widens the application range.

The present disclosure relates to the technical field of computer vision, and proposes a neural radiance field based open set semantic segmentation method and system, which comprises the following steps: performing feature extraction on the obtained to-be-recognized image based on neural radiance field to obtain global perspective features of known class objects; calculating the hybrid distance between the global perspective features of the known class objects and the corresponding prototypes to constrain all known class features to be close to the corresponding class prototypes in the metric space, and obtaining a learned semantic field; and performing segmentation on unknown objects in a scene according to the learned semantic field by using a maximum Logits strategy based on metrics combined with an open operation method of morphology. The present disclosure uses an implicit neural scene expression method for open set segmentation, improves the feature expression capability and open set segmentation performance, and can realize accurate segmentation of unknown class objects in a scene while maintaining accurate segmentation of known class objects in the scene.

A sparse projection CBCT reconstruction method, device and equipment and readable storage medium, which comprises the following steps: obtaining X-ray projections of a target patient at multiple angles; obtaining a whole tissue mask prediction projection for reconstruction and a hard tissue mask prediction projection for reconstruction according to a preset tissue projection prediction model and the X-ray projections; supervising and training a preset neural field network initial model according to the X-ray projections, the whole tissue mask prediction projection for reconstruction and the hard tissue mask prediction projection for reconstruction to obtain a neural field network model corresponding to the target patient; obtaining whole tissue gray scale data of each coordinate point in a three-dimensional image to be reconstructed according to the neural field network model and coordinate information of each coordinate point in the three-dimensional image to be reconstructed, and forming the three-dimensional image to be reconstructed. Different tissues in CBCT are modeled by a multi-unit network, the interpretability and controllability of the neural field network model are improved, and the method has strong practical significance.

The invention provides a method and a system for identifying a connecting line relationship between primitives, which comprises the following steps of: constructing a primitive connecting line ambiguity perception model, and performing model pre-training on the primitive connecting line ambiguity perception model based on adversarial training and a multi-task joint loss function, obtaining an original primitive wiring diagram, inputting the original primitive wiring diagram into the primitive connecting line ambiguity perception model for ambiguity perception, generating a connecting line ambiguity perception knowledge graph, inputting the connecting line ambiguity perception knowledge graph into a pre-constructed circuit neural field for analogue simulation and decision screening, generating a stubborn connecting line ambiguity perception result, and outputting the stubborn connecting line ambiguity perception result to the primitive connecting line ambiguity perception result. And S4, analyzing the stubborn connection ambiguity perception result based on a stubborn connection ambiguity analysis mechanism, generating the stubborn connection ambiguity result and expanding the stubborn connection ambiguity result to a connection ambiguity perception knowledge graph, iteratively executing the identification processes of the steps S3 to S4 until a preset condition is met, ending iterative identification, and outputting a primitive connection relation identification result. Therefore, accurate and efficient identification of the circuit connection relation is realized.

The invention relates to the technical field of computer graphics and artificial intelligence, and discloses a mobile terminal-oriented lightweight 3D scenic spot model dynamic generation and rendering method, which comprises the following steps of: firstly, constructing a global material primitive library at a server side, and parameterizing each primitive into a material space coordinate vector; then, a neural field function is independently trained for the scene partition, and the function learns to map space coordinates to field density and a material mixed vector; on the mobile terminal, according to a viewpoint dynamic loading function, generating a geometric surface in real time by using field density through a volume rendering principle; and meanwhile, performing proximity search and weighted interpolation in the material primitive library by using the material mixed vector, and synthesizing a high-precision surface material in real time. According to the invention, huge geometric and texture data are implicitly coded into a lightweight network, so that the data storage and transmission pressure is remarkably reduced, and high-fidelity real-time rendering on performance-limited equipment is realized.

The invention discloses a cross-platform virtual reality content sharing and synchronizing method, and relates to the technical field of virtual reality cooperative computing, and the method comprises the steps: collecting real-time interaction data, generating a dynamic neural field feature tensor, and detecting abrupt change nodes to construct a difference data set; converting the difference data set into light field tokens, and distributing the light field tokens to heterogeneous terminals; analyzing the light field token in the heterogeneous terminal to restore the high-dimensional curvature feature vector, and generating a conflict object list through a space-time convolution conflict detection model; based on the conflict object list, conflict resolution is executed through a quantum hierarchical arbitration decision, and an arbitration quantum state vector is generated; driving photon field rendering according to the arbitration quantum state vector, and outputting a rendering result through a terminal strategy and optical projection; the photon field rendering is driven according to the arbitration quantum state vector, and the consistency of the visual experience of the heterogeneous terminal is guaranteed.

The invention discloses an all-weather scene adaptive reconstruction method based on non-uniform point cloud and depth map fusion, and the method comprises the steps: collecting multi-modal sensor data, carrying out the time-space alignment, and unifying the data to a world coordinate system; performing multi-scale sparse Hash coding on the data after space-time alignment to generate high-dimensional feature representation of the query point; implicit neural scene representation decoding based on weather perception: inputting the high-dimensional feature representation and the weather condition vector into a geometric decoder, and outputting a symbol distance function value and a geometric feature vector; gradient-level adaptive fusion is realized by using a dynamic prior weight network, a dynamic weight is generated according to a spatial position, an environmental condition and local uncertainty, and gradient back propagation in a training process is modulated; and obtaining a finally reconstructed three-dimensional scene model through end-to-end joint training and scene reconstruction. According to the method, high-precision and high-robustness unified reconstruction of non-uniform point cloud and depth map data under extreme weather interference is realized.

The application provides a 4D space-time field construction method and device based on neural field reconstruction, and relates to the technical field of data processing. The 4D space-time field construction method based on neural field reconstruction comprises the following steps: performing supervision information labeling on a plurality of three-dimensional coordinates in a to-be-processed 3D model to obtain a plurality of training data samples, thereby constructing a training data set; dividing a three-dimensional space into a plurality of cubic grids of the same size, each grid corresponding to a neural network inside, thereby constructing a block neural field model; training the block neural field model based on the training data set to obtain a target 4D space-time field; and updating the target 4D space-time field in response to local data update of the to-be-processed 3D model. Through block neural field modeling and a continuous learning mechanism, the application converts dynamic update of the 4D space-time field into an incremental inference problem of the neural network, thereby realizing efficient update of the 4D space-time field.

This invention proposes an integer-order differential analog-to-digital converter (ADC) for intracranial neural potential acquisition. First, an integer-order differential operator is designed to calculate arbitrary integer-order incremental codes, which more effectively utilizes the sparsity of intracranial neural potentials, significantly improving the compression rate of intracranial neural potentials. It also offers greater flexibility, with adjustable differential order, adapting to a wider range of intracranial neural potential acquisition applications. Second, a multi-level buffer matrix module is designed to generate arbitrary integer-order incremental code values ​​in a concise and efficient manner. Finally, a serialized bitstream generator is designed to serialize and package integer-order incremental data acquired from multiple parallel channels. This not only offers high scalability but also allows for reliable interfacing with subsequent synchronous clock wireless transmission systems, avoiding the overhead and latency of synchronous-asynchronous interface coordination, further improving signal acquisition efficiency.

The invention provides a 4D space-time field construction method and device based on neural field reconstruction, and relates to the technical field of data processing. A 4D space-time field construction method based on neural field reconstruction comprises the following steps: performing supervision information labeling on a plurality of three-dimensional coordinates in a to-be-processed 3D model to obtain a plurality of training data samples so as to construct a training data set; the three-dimensional space is divided into a plurality of cubic grids of the same size, a neural network corresponds to the interior of each grid, and therefore a block neural field model is constructed; training a block neural field model based on the training data set to obtain a target 4D space-time field; and in response to local data updating of the 3D model to be processed, updating the target 4D space-time field. According to the method, dynamic updating of the 4D space-time field is converted into an incremental reasoning problem of a neural network through block neural field modeling and a continuous learning mechanism, and efficient updating of the 4D space-time field is achieved.

The invention discloses a multi-modal-based wind power plant digital twinborn interaction method and device and a storage medium, and the method comprises the steps: collecting a real-time image of a wind turbine generator, completing the calibration of an image pose, forming a three-dimensional scene with a NeRF voxel grid, importing CFD flow field simulation data, mapping a flow field scalar value to the NeRF voxel grid, obtaining an enhanced neural scene, and carrying out the simulation of the flow field. The global continuity of simulation data is considered at low wind speed, wake flow distribution of the whole wind power plant can be clearly seen, local turbulence of a single unit can be accurately positioned, a multi-modal model is deployed, professional terms in the field of wind power operation and maintenance are preset, after an instruction is sent, the model combines state information of an enhanced neural scene, the instruction is analyzed into a recognizable sequence, and the recognition efficiency of the wind power operation and maintenance is improved. The geometric structure and physical field distribution of a neural scene are enhanced through sequence combination, an optimal observation path is generated, the track smoothness is adjusted according to the eye movement angular velocity of a user, the correction amplitude is amplified during high eye movement, and view angle switching lagging is avoided through focusing point anchoring.

A method and apparatus for dataset distillation are provided. The method includes: obtaining an original coordinate set including coordinates of original data included in an original dataset; generating first test data included in a test dataset corresponding to a first neural field model from among a plurality of neural field models by providing the original coordinate set to the first neural field model; obtaining a first result by providing at least a portion of the original dataset to a neural test model; obtaining a second result by providing at least a portion of the test dataset including the first test data to the neural test model; determining a distillation loss based on the first result and the second result; and training the plurality of neural field models based on the distillation loss.

The invention provides a model resource processing method and device, a computer program product and electronic equipment. The model resource processing method comprises the following steps: determining a geometric grid corresponding to a target neural field model based on the target neural field model and original images under a plurality of observation view angles; performing UV expansion on a geometric grid corresponding to the target neural field model to obtain a UV layout of the geometric grid, and performing reverse solution based on the UV layout of the geometric grid to obtain a PBR material map of the geometric grid; semantic instance segmentation information of each original image is determined through semantic instance segmentation, and the semantic instance segmentation information of each original image is projected to a surface corresponding to the geometric grid to obtain a semantic instance label corresponding to the geometric grid; and packaging the geometric grid, the PBR material map of the geometric grid and the semantic instance label corresponding to the geometric grid into a standardized PBR asset package based on a preset packaging format so as to improve the reusability of model resources.

The application discloses a virtual air station quality control data generation method based on a space-time Bayesian neural field, relates to the technical field of data processing, and can capture and enhance long-range space-time dynamics in discontinuous observation data, directly generate virtual station air quality results, namely, a target pollutant concentration set, under various missing data modes, and provide a confidence interval of a corresponding space-time point, so that the accuracy of air station quality control data can be improved.