16018 results about "Algorithm" patented technology

The invention discloses an image processing method and device and a storage medium. The method comprises the steps of obtaining a to-be-simulated 3D convolution model and training data; decomposing the 3D convolution model into cascading of a 3D space convolution model and a 3D time convolution model to obtain a pseudo 3D cascading convolution model; training a pseudo 3D cascade convolution modelby using the training data, and obtaining parameters of a 3D spatial convolution model and a 3D time convolution model; converting the 3D space convolution model and the 3D time convolution model intoa 2D space convolution model and a 2D time convolution model; setting a feature rearrangement rule for the 2D spatial convolution model and the 2D time convolution model; mapping model parameters ofthe 3D spatial convolution model and the 3D time convolution model into parameters of a 2D spatial convolution model and a 2D time convolution model to obtain a 2D cascaded convolution model; and performing convolution operation on the image by using the 2D spatial convolution model and the 2D time convolution model. By means of the mode, image processing conducted through 3D convolution operationcan be achieved through the 2D convolution model.

A second level intra prediction mode can be combined with one or more of sixty-seven JVET intra prediction modes during encoding of a coding unit in a video bitstream. Embodiments include making a position dependent intra prediction combination (PDPC) mode available as the second level intra prediction mode. In embodiments, when a PDPC (position dependent intra prediction combination) mode is enabled, the second level intra prediction is combined with one of the 67 selected intra predictor modes. In embodiments, the PDPC mode is only enabled or available for a predetermined subset of intra prediction modes (out of 67 possible modes), in order to reduce encoder complexity and potentially improve coding efficiency. The PDPC mode may be identifies as enabled or available by a list of modes or signaling in the video bitstream.

Devices and techniques are generally described for nonlinear tensor compression for neural networks. In various examples, a first tensor associated with a first layer of a neural network may be determined. One or more neural processing units of accelerator hardware may generate a first compressed tensor by applying a nonlinear compression function to the first tensor. The first compressed tensor may be stored in a first memory of the one or more computer-readable media. A first operation associated with a second layer of the neural network may be determined, where the first operation uses output of the first layer. The first operation may be performed based on the first compressed tensor.

A radio frequency (RF) signal classification device may include an RF receiver configured to receive RF signals, a quantum computing circuit configured to perform quantum subset summing, and a processor. The processor may be configured to generate a game theory reward matrix for a plurality of different deep learning models, cooperate with the quantum computing circuit to perform quantum subset summing of the game theory reward matrix, select a deep learning model from the plurality thereof based upon the quantum subset summing of the game theory reward matrix, and process the RF signals using the selected deep learning model for RF signal classification.

Examples relate to a concept for an entry-exit matching system, and in particular to an evaluation device, a method and a computer program for person re-identification for entry-exit matching in a transportation system. The evaluation device comprises processing circuitry configured to obtain a plurality of re-identification codes. Each re-identification code represents a person being recorded by at least one camera when entering or exiting at least a section of the transportation system. The processing circuitry is configured to match the plurality of re-identification codes using a global matching scheme to obtain a plurality of matched pairs of re-identification codes, such that each matched pair of re-identification codes comprises a re-identification code of a person entering and a re-identification code of a person exiting. The global matching scheme is based on reducing an overall distance between the re-identification codes of the matched pairs of re-identification codes over the plurality of matched pairs of re-identification codes. The processing circuitry is configured to determine points of entry and exit for the plurality of matched pairs of re-identification codes.

This invention discloses a stepwise data assimilation method for a set subspace in a nonlinear inverse problem, relating to the field of data processing technology. The invention constructs an initial prior physical set based on the physical state variables to be inverted and optimized, and historical observation data. It extracts the static subspace basis anomaly matrix, initializes the latent variable set and particle weights, calculates fractional-step incremental reweighting, evaluates the current likelihood mismatch penalty using predicted data, updates and normalizes the particle weights in the logarithmic domain, obtains the effective sample number, performs system resampling operations in conjunction with a preset resampling tolerance coefficient, eliminates low-weight particles and replicates high-weight particles, synchronously updates the latent variable set and predicted data, executes a dynamic MCMC mutation loop to generate proposed latent state vectors and affinely maps them to a high-dimensional physical space, and updates the latent variable set by evaluating the annealing target energy within the latent variable subspace until all fractional steps are traversed. Finally, it outputs the latent variable set and maps it back to the physical space.

The various implementations described herein include methods and systems for encoding video. In one aspect, a method includes receiving a mesh with polygons representing a surface of an object; detecting a first symmetric region in the mesh that includes a first symmetry line to divide the first symmetric region into a first partition and a second partition. The method includes recursively determining whether one of the first partition or the second partition includes a second symmetric region until no symmetric region is detected in both the first partition and the second partition. The method includes in response to detecting the second symmetric region within one of the first partition or the second partition: determining a second symmetry line within the second symmetric region to divide the first or the second partition into a third sub-partition and a fourth sub-partition; and compressing information of the third sub-partition, the second symmetry line and the first symmetry line into a bitstream.

Systems and methods are provided for analog hardware realization of neural networks. The method incudes obtaining a neural network topology and weights of a trained neural network. The method also includes transforming the neural network topology into an equivalent analog network of analog components. The method also includes computing a weight matrix for the equivalent analog network based on the weights of the trained neural network. Each element of the weight matrix represents a respective connection between analog components of the equivalent analog network. The method also includes generating a schematic model for implementing the equivalent analog network based on the weight matrix, including selecting component parameter values for the analog components.

A first large language model (LLM) instance may be instructed to request data while being prevented from performing calculations using the data. A second LLM instance may be instructed to provide a response to the request for data based on a known complete data set. The response may be translated into a machine-readable response in a format configured for processing by a calculation engine. The calculation engine may process the machine-readable response, thereby generating a calculation engine output. A mismatch between the calculation engine output and a known result obtained using the known complete data set may be identified, and the instruction to the first LLM may be modified in response.

Some implementations relate to methods, systems, and computer-readable media for adapting simulated character interactions to different morphologies and interaction scenarios. The system accesses a graph representing a control policy for a simulated character's movements in a virtual environment. This graph undergoes encoding and processing through a graph neural network to generate latent embeddings for the graph. A fixed-length latent vector is determined from the latent embeddings. This vector is input to a feedforward neural network, generating control signals for the character's actions. Through a reinforcement learning loop, the character's motions are continuously refined by iteratively adjusting the graph based on evaluating the actions of the simulated character via a reward function, adapting the control policy to different character morphologies and / or interaction scenarios.

A size of a transform block is identified. A transform type for the transform block is identified. The transform type includes a horizontal transform type and a vertical transform type. Identifying the transform type includes determining whether the size of the transform block is below a predefined block size; and, in response to determining that the size is below the predefined block size, selecting a default transform type for each of the horizontal transform type and the vertical transform type. The transform type is then applied to the transform block. The default transform type can be the discrete cosine transform (DCT).

The application discloses an infrared long-distance space adjacent target super-resolution method, wherein the method comprises the following steps: S1, analyzing the imaging characteristics of a long-distance target optical detection system and the imaging characteristics of space adjacent multiple targets, modeling point target imaging, and generating a simulation image dataset; S2, initializing the input simulation image based on linear mapping; S3, realizing infrared space adjacent target super-resolution modeling by adopting a sparse reconstruction algorithm, and constructing a deep unfolding network framework; S4, sequentially passing the initialized image through each stage of the deep unfolding network; S5, constructing a GPU training environment, setting a data loader, a model and an evaluator configuration file, training the model, applying the trained model to a test set, generating a high-resolution image, extracting target coordinate information by post-processing and inputting the target coordinate information into the evaluator, obtaining evaluation indexes based on infrared long-distance space adjacent targets, and calculating an average detection rate.

Techniques for training and using machine learning models for sensitivity detection. A method for sensitivity detection training includes fine-tuning a language model by iteratively applying the language model to prompts and adjusting weights of the language model. The prompts indicate classifications for a set of first resources and characteristics of an entity. The fine-tuned language model is queried with respect to classifications of a set of second resources. The fine-tuned language model is queried using prompts indicating the second classifications and data indicating characteristics of an entity, where outputs of the language model include a sensitivity for each of the second classifications. Training data including the second classifications is labeled based on the sensitivities output by the language model. A sensitivity detection machine learning model is trained using the labeled training data set such that the trained sensitivity detection machine learning model is configured to output sensitivities for resource classifications.

A time series of raw performance data samples pertaining to performing of a set of repetitions of a movement by a user is collected from a sensor. A set of features is generated from the collected time series of raw performance data samples, including by extracting one or more waveform shape features. The set of features, including the extracted one or more waveform shape features, is provided as input to a model that outputs an estimate of repetitions in reserve.

The present application relates to the technical field of geological disaster risk assessment, and discloses a landslide-prone prediction method based on multi-scale geographic weighted regression and spatial heterogeneity partitioning, which solves the problems of low local accuracy, poor area efficiency and insufficient prediction stability caused by single spatial scale, uniform modeling and single evaluation system in traditional landslide prediction models. The present application organically combines multi-scale geographic weighted regression with spatial heterogeneity partitioning, first adaptively assigns a dedicated optimal bandwidth to each disaster-inducing factor through a multi-scale regression model to accurately depict the spatial non-stationary characteristics of landslide disaster-inducing relationships, then reconstructs the discrete regression coefficients into a spatial continuous surface with the help of Kriging interpolation, dynamically extracts the dominant factors with the optimal discrete and natural discontinuity method to realize the heterogeneity partitioning of the study area, takes each partition as an independent unit to carry out landslide-prone deduction, and constructs a multi-dimensional comprehensive evaluation system from the dimensions of prediction accuracy, area rationality and uncertainty robustness.

Methods for video decoding and encoding, apparatuses, and non-transitory computer-readable storage media are provided. In one method, a decoder may obtain a first block vector (BV) and a second BV based on a bi-predicted intra block copy (IBC) mode. Additionally, the decoder may generate a first prediction block based on the first BV and generating, by the decoder, a second prediction block based on the second BV. Furthermore, the decoder may obtain a final prediction block for a current block based on the first prediction block and the second prediction block.

Implementing a neural network in hardware, particularly as an accelerator. The neural network comprises a plurality of layers and the layers are grouped into a plurality of layer groups, each layer gr

A coder obtains, for a block of video, first counts of occurrences of angular intra prediction modes (IPMs) used by neighboring blocks of the block. The coder obtains, for the block, second counts of occurrences of non-angular intra coding modes used by the neighboring blocks. Based on the first counts, a plurality of angular IPMs is determined from the angular IPMs. Based on the second counts, whether to enable non-angular coding mode fusion with the plurality of angular IPMs is determined. The coder codes the block using a prediction block generated based on the determining of whether non-angular intra coding mode fusion is enabled. For an encoder, the block is coded by signaling a residual block as a difference between the block and the prediction block. For a decoder, the block is coded by combining the prediction block with the residual block to reconstruct the block.

Provided are a frequency-domain data merging method and apparatus, a storage medium, and an electronic apparatus. The method includes: respectively performing up-conversion on baseband data of a plurality of frequency bands in a frequency domain, so as to obtain frequency-domain received data of each frequency band from among the plurality of frequency bands that has been subjected to up-conversion; and merging the frequency-domain received data of each frequency band that has been subjected to up-conversion, so as to obtain merged frequency-domain received data.

1. Name of the product in this design: High-definition data cable (round tail design 2). 2. Purpose of this product design: High-definition data cable for data transmission. 3. The key design feature of this product is its shape. 4. The image or photograph that best illustrates the design's key points: a 3D model.

This application provides a method for optimizing estimation parameters, a controller, a vehicle, and a storage medium. The method includes: acquiring multiple residual values ​​of the vehicle within a first preset time interval, where the residual values ​​are the differences between predicted and measured values ​​of the vehicle's motion state parameters; determining the uncertainty level of the multiple residual values ​​based on their probability values ​​within each preset residual value interval; determining a scaling factor for initial estimation parameters based on the uncertainty level of the multiple residual values; and optimizing the initial estimation parameters based on the scaling factor to obtain optimized estimation parameters. By determining the scaling factor of the initial estimation parameters based on the uncertainty level, the vehicle can adaptively adjust the initial estimation parameters according to the operating conditions to obtain optimized estimation parameters, improving the scenario adaptability of the estimation parameters, thereby improving the estimation accuracy and stability of the vehicle's dynamic parameters and reducing the cost caused by repeated manual parameter calibration.

This application discloses a testing method and related apparatus for an AI all-in-one machine, relating to the field of artificial intelligence. The method includes: identifying the hardware combination information of the AI ​​all-in-one machine, including the types of multiple AI cards and their operating environment information; generating a hardware adapter module matching the types of the AI ​​cards, the hardware adapter module encapsulating a basic computing power testing interface; running a containerized testing environment matching the operating environment information of the multiple AI cards; in the containerized testing environment, testing the floating-point computing performance and data transmission performance of a single AI card in the AI ​​all-in-one machine, as well as the collaborative processing performance among multiple AI cards in the AI ​​all-in-one machine, through the basic computing power testing interface provided by the hardware adapter module; and generating basic computing power test results for the AI ​​all-in-one machine based on the test results of the floating-point computing performance, data transmission performance, and collaborative processing performance. This application effectively realizes the basic computing power testing of AI all-in-one machines.

This invention belongs to the fields of rock mechanics, signal processing, and deep learning. It provides an intelligent method, device, equipment, and medium for identifying rock fracture types. The method includes: acquiring acoustic emission signal data, loading stress-strain data, and rock image / video of the rock sample; extracting multimodal features from the acoustic emission signal data and loading stress-strain data; performing principal component analysis on the multimodal features to obtain a dimensionality-reduced feature matrix; labeling training data using a rock failure event time-series correlation method to obtain crack type labels; training and detecting using a gated recurrent unit network based on the dimensionality-reduced feature matrix and crack type labels to obtain a shear crack probability sequence, and then calculating the tension-shear ratio within a time step; and generating time stamps in the rock image / video based on the tension-shear ratio. This invention achieves accurate and rapid identification of rock fracture types and real-time monitoring of the dynamic evolution of fractures.

The application relates to a kind of body image matching network assisted internal three-dimensional deformation field measurement method, belong to deformation measurement field.The application implementation method is: in reference body image, define calculation region and calculation voxel point;Using three-dimensional body image feature extraction network, key three-dimensional features are extracted simultaneously.Using three-dimensional feature matching network, key three-dimensional features are matched.From matched three-dimensional feature point pair, the feature point pair in the preset three-dimensional neighborhood range around each calculation voxel point is extracted, and the three-dimensional deformation initial value of the calculation voxel point is calculated when more than four groups of feature point pairs are judged;Less than 4 groups are directly assigned using the deformation initial value of other successfully calculated points with the spatial position of the calculation voxel point nearest to the interpolation or, as the three-dimensional deformation initial value of the point.Using three-dimensional reverse combination Gauss-Newton nonlinear iterative optimization method, high-precision internal three-dimensional deformation field measurement value is calculated, that is, internal three-dimensional deformation field measurement is realized.

Systems, methods, and instrumentalities are configured for intra template matching prediction fusion. In examples, a video decoding device may be configured to determine that intra template matching prediction (ITMP) fusion is used for a video block. The device may obtain a selected block vector (BV) indication configured to indicate a selected BV candidate. The device may determine, based on the selected BV candidate, a set of BV candidates for fusion. The device may reconstruct the video block based on the determined set of BV candidates.

This application discloses an industrial inspection method, apparatus, equipment, product, and storage medium, relating to the field of industrial vision inspection technology. The method includes: responding to an industrial inspection command, acquiring an image to be inspected for membrane defects; processing the image using a preset defect detection model to obtain a defect detection result. The preset defect detection model is obtained by quantizing a target defect detection model, which is obtained by structural pruning and lightweight reconstruction of an initial defect detection model. The structural pruning and lightweight reconstruction are used to reduce the number of model parameters and computational load. This application reduces the number of model parameters and computational load through structured pruning and lightweight reconstruction, and then quantizes the model, enabling the quantized model to be deployed on edge devices while improving the model's inference efficiency, achieving efficient model compression and real-time inference.

This application discloses a method, apparatus, storage medium, and computer device for visual environment detection. The method includes: in response to receiving twin model update information corresponding to a target park, generating a content area distribution map corresponding to each content type based on key events within a target time period, the frequency of occurrence of key events, the number of times the digital twin model corresponding to each sub-area is displayed, and key content tags under the content type; generating a prediction model corresponding to each area based on the different importance of each area in the content area distribution map, wherein each prediction model shares a common model module; determining a sequence of visual environment detection distribution maps under the content type corresponding to the content area distribution map using the prediction model set; generating a corresponding twin model based on the sequence of visual environment detection distribution maps; and in response to receiving twin model display information for at least one content type, overlaying and displaying the corresponding at least one twin model.