230results about How to "Efficient compression" patented technology

The invention relates to integrated airborne communication antenna equipment and a communication method. The method comprises the following steps: carrying out CSI data acquisition and space-time focus modulation on a communication channel to obtain CSI space-time characteristic data; performing residual vector quantization operation to generate a quantized data stream, and performing recursive coding processing to obtain a target data stream; mapping the target data stream to a three-dimensional tensor space, constructing a real value tensor model, and performing iterative operation to obtain a CSI feature vector; performing fractional order wavelet transform decomposition on the CSI feature vector to obtain a radio frequency feature set, and inputting the radio frequency feature set into a federal fractional scattering network for feature processing to obtain an identity authentication result; and calculating an amplitude-phase weight coefficient of the antenna array, generating a beam parameter, and performing compensation matrix operation on the beam parameter according to the CSI feature vector to obtain a pre-coded signal. According to the invention, the dynamic adjustment of the antenna beam parameters is realized, and the accuracy and anti-interference capability of beam forming are improved.

The invention relates to the technical field of data storage, and discloses a lossless compression algorithm-based data storage method, which comprises the following steps of: receiving to-be-stored data through a data interface, automatically identifying a data format by using a data analysis tool, and performing feature analysis to obtain the size, the type and the complexity of the data; through deep analysis of data features, the method can flexibly select a proper Huffman coding processing mode, and optimizes a compression process according to different data types and complexity. In addition, a context modeling optimization mechanism is introduced, the coding mode is dynamically adjusted, and the compression ratio and the storage efficiency are further improved; efficient storage and reliability verification of data are ensured through a block storage strategy in combination with metadata management; the problem that an existing compression algorithm is low in efficiency is solved, the integrity and reliability of data storage are improved, and the method is suitable for storage requirements of large-scale data and diversified data.

The invention relates to the technical field of text processing, in particular to a text compression method, a text decompression method, a model training method, devices and equipment.The text compression method comprises the steps that a to-be-compressed target text is obtained, and the to-be-compressed target text is input into a feature extraction module of a text compression model; using a deep learning technology to capture semantic and syntactic features of the text, and expressing the extracted key information in the form of an embedded vector sequence; and a q-former quantization processing module is used for carrying out quantization processing on the embedded vector sequence, and the continuous embedded vector sequence is converted into a discrete token sequence, so that effective compression of text information is realized. By adopting a form of combining a feature extraction module and a q-former structure, the redundancy of text data can be remarkably reduced, and meanwhile, the high-quality restoration capability of an original text is kept.

The invention discloses a radar and video information fusion coding method for vehicle over-limit early warning. The method comprises the following steps: acquiring three-dimensional point cloud data through a radar module; measuring three-dimensional sizes of length, width and height of the target vehicle in real time; a video information acquisition module is used for acquiring vehicle image data, and the video data identifies license plate numbers and vehicle appearance features through an improved YOLO deep learning algorithm to assist in judging whether a vehicle exceeds a limit or not; the vehicle three-dimensional size measured by the radar is compared with the standard size in the database through the data processing module, whether the vehicle exceeds the limit or not is judged through a threshold value method, radar data and video data are effectively fused through an algorithm, and the fused data are compressed. The radar and video information fusion technology is utilized, an efficient coding scheme is combined, the high-precision and high-efficiency vehicle over-limit early warning function can be achieved, and the method is suitable for complex road environments and variable traffic scenes.

The invention provides a sparse representation-based low-bit-rate video coding method, which comprises the following steps of: after a coding end acquires an input video, carrying out frame segmentation sampling on the input video to obtain a video frame sequence, and carrying out block segmentation on elements in the video frame sequence to obtain an image block sequence; processing elements in the image block sequence based on a dictionary learning algorithm and a sparse representation algorithm of gradient sparsity to obtain a sparse representation coefficient; sequentially performing quantization processing and entropy coding processing on the sparse representation system to obtain a compressed bit stream, and storing and transmitting the compressed bit stream; obtaining a compressed bit stream, and sequentially performing entropy decoding processing and inverse quantization processing on the bit stream to obtain an original sparse representation coefficient; processing the original sparse representation coefficient based on a dictionary learning algorithm and a sparse decoding algorithm of gradient sparsity to obtain a reconstructed image block; merging the plurality of reconstructed image blocks to obtain a reconstructed video frame; and carrying out frame combination on the plurality of reconstructed video frames to obtain a reconstructed input video.

The invention provides an image classification method and device based on a student model, equipment and a medium, and relates to the technical field of knowledge distillation, and the method comprises the steps: obtaining a to-be-classified image; inputting the to-be-classified image into a target student model to obtain a category corresponding to the to-be-classified image; inputting the initial image set into a trained teacher model and a preset initial student model, and obtaining a teacher model imbalance category probability and a student model imbalance category probability; inputting the mixed image set into a trained teacher model and a preset initial student model, and obtaining a teacher model imbalance category probability and a student model imbalance category probability; determining a corresponding loss; according to the loss, performing back propagation to update student model parameters so as to obtain a target student model; when the image is classified, the resource computing power required by computer hardware can be reduced, the response speed of a computer is improved, and the phenomenon of computer jamming is avoided.

The invention discloses a drug target combinability prediction method based on variational coding, and belongs to the field of bioinformatics. According to the method, efficient extraction of drug and target features, construction of potential feature space of the drug and target features and prediction of the interaction relationship of the drug and target features are realized through a dual-stage coupled end-to-end neural network. The model is mainly composed of two stages, in the first stage, potential representations of drug and target features are obtained through variational compression and feature extraction, and in the second stage, a drug target action matrix is reconstructed through an up-down sampling comprehensive deep neural network. Through the architecture, the chemical characteristics of the drug and the biological characteristics of the target can be effectively integrated, and the accuracy of interactive relationship prediction is improved. The invention provides a novel and efficient solution for deep analysis of biological big data at the same time.

The invention relates to the field of shower equipment, and provides a foaming shower which comprises a shell, an electromagnetic valve, a plurality of liquid storage tanks, a liquid pump, a clear water inlet assembly, a function control assembly, a foaming assembly and a water outlet assembly. The function control assembly comprises a water inlet shell, a function control valve element and a plurality of contact induction devices. The output end of the clear water inlet assembly is connected with the input end of the water inlet shell, the clear water output end of the water inlet shell is connected with the clear water port of the water outlet assembly, and the foam water output end of the water inlet shell is connected with the water inlet port of the foaming assembly; the function control valve element is provided with a trigger block, the function control valve element drives the trigger block to rotate and can trigger the contact induction devices, and the contact induction devices are connected with the electromagnetic valves in an electric control mode. The plurality of liquid storage tanks are respectively connected with the input end of the electromagnetic valve; the problems that an existing shower is limited in foam type and number, foam remains when clear water is converted, and the use experience is poor can be effectively solved.

The invention discloses a high dynamic range image tone mapping method and device, a medium and a program product, and relates to the field of image processing. The method comprises the following steps: calculating a brightness logarithm and a large-scale texture layer logarithm of a high dynamic range image to enhance image details and generate a global detail enhanced image; the brightness distribution is optimized by counting the brightness histogram, cutting the abnormal brightness value and compensating the abnormal brightness value; determining the average brightness and the standard deviation of the low dynamic range image based on the optimized histogram, obtaining a normalized brightness grey-scale map, and performing local detail enhancement; and restoring the brightness of the enhanced image to an original HDR range, mapping a color channel, and merging color spaces to obtain a final LDR image. According to the method, the dynamic range of the high-dynamic-range image can be compressed, and meanwhile, the detail features and brightness distribution reasonability of the image are effectively kept, so that the subjective visual quality and user experience of the low-dynamic-range image are remarkably improved.

The invention relates to a model processing method and device, computer equipment, a readable storage medium and a program product. The method comprises the steps of performing low-dimensional conversion processing on input data, a target layer and an activation function of a next layer of the target layer based on a diffusion accounting method to obtain a converted feature vector; based on a preset interlayer similarity algorithm and the converted feature vector, constructing a similarity matrix between the target layer and the next layer of the target layer; and determining similar layers in the similarity matrix, fusing the similar layers according to preset fusion parameters to obtain fused similar layers, and obtaining a compressed large language model based on the fused similar layers. By adopting the method, the model scale can be effectively compressed while the performance of the large language model is ensured.

The invention discloses an intelligent data compression method for a twin system. The method comprises the following steps: collecting video, audio and text data for preprocessing; constructing a Bootstrap model, and performing feature extraction and preliminary compression on the input data; the method comprises the steps that a Combined model composed of a Bootstrap model and a Support model is constructed, and probability distribution output by the Bootstrap model is combined with probability distribution output by the Support model; finally compressing the output probability distribution by using an arithmetic coding method so as to reduce the volume of data transmission, and transmitting the data to target equipment; the target device decodes and reconstructs the received data and restores the original data before compression; according to the method, the data compression ratio and the compression time efficiency are improved, so that the time delay of data transmission is effectively reduced, the method is suitable for various data types in a twin system, and real-time processing of the data is ensured.

The invention discloses a CSI (Channel State Information) feedback method and system based on deep learning, belongs to the technical field of communication, and aims to solve the technical problems of low recovery precision, reduced frequency efficiency and huge uplink bandwidth consumption of CSI feedback in a large-scale MIMO (Multiple Input Multiple Output) system. Comprising the following steps that a channel state information feedback model TCMRNet is constructed and trained, the TCMRNet comprises an encoder and a decoder, LSTM network units and convolution layers with different convolution kernel sizes are introduced into the encoder and the decoder, the output end of each convolution layer is connected with a batch normalization layer, and the output end of each convolution layer is connected with an LSTM network unit; the encoder is used for carrying out multi-resolution feature extraction and time correlation analysis on input channel state information and compressing extracted features into feature vectors to be output, and the decoder takes the feature vectors as input and is used for decompressing the feature vectors, recovering the channel state information and outputting reconstructed channel state information; and performing time domain transformation on the reconstructed channel state information.

The invention discloses a deep network model quantization compression method, which relates to the technical field of model quantization and comprises the following steps of: preprocessing a deep network model, compressing the deep network model by combining model quantization and pruning, performing low-rank decomposition on the deep network model and verifying a compression quantization effect of the deep network model; according to the method, a series of preprocessing operations are carried out on the deep network model to be processed, the deep network model can be ensured to be smoothly carried out in a subsequent quantization process, the performance of the model is reserved as far as possible, and redundant parameters in the deep network model can be deleted by carrying out model pruning on the deep network model in combination with model quantization. According to the method, the complexity of the model can be reduced, the precision of the model can be kept, floating-point number parameters in the model are converted into low-precision fixed-point numbers, and the storage and calculation cost of the model can be reduced, so that the deep network model is effectively compressed and quantized, and the deep network model is lighter.

The invention discloses a waste mulching film recycling device, and relates to the technical field of agricultural equipment, the waste mulching film recycling device comprises a traction frame, the traction frame is fixedly connected with a scraper knife, the left and right ends of the traction frame are rotatably connected with rollers, the top of the traction frame is fixedly connected with a transmission case, the transmission case is connected with an output shaft, and the output shaft is rotatably connected with a support; the output shaft is fixedly connected with the circle centers of the belt wheels, the two belt wheels are in transmission connection through a belt, the belt wheels are rotationally connected with the side wall of the traction frame, and a transmission motor drives an incomplete gear to be matched with an auger, so that efficient compression and forward conveying of the waste mulching film are achieved, and the overall size of the waste mulching film is reduced; quantitative cutting of the mulching film is achieved, follow-up recycling of the stacked mulching film is facilitated, the follow-up treatment process can be optimized, the overall treatment efficiency is improved, and continuous operation of recycling of the mulching film is facilitated.

The invention discloses a tool wear state identification method and system for cutting of a numerical control machine tool. The method comprises the following steps: acquiring a three-way vibration signal, a power signal and a temperature signal; performing time domain, frequency domain and time-frequency domain multi-domain feature extraction on the three-way vibration signal, the power signal and the temperature signal, and constructing a multi-domain high-dimensional feature vector; performing feature fusion dimension reduction on the high-dimensional feature vector to obtain a feature sample set; obtaining a cutter state recognition model according to the feature sample set; and inputting the acquired multi-source signals into a cutter state identification model to carry out cutter wear state identification of cutting machining of the numerical control machine tool. By the adoption of the technical scheme, various machining conditions and tool types can be better adapted, and the applicability and universality of the recognition system are improved.

The utility model relates to the field of industrial waste gas, and discloses a waste gas sample storage device for industrial waste gas detection, which comprises a gas storage cylinder, an exhaust hole formed in the center of the bottom of the gas storage cylinder, an insertion hole formed in the center of the top of the gas storage cylinder, a sealing sleeve fixedly connected to the inner side wall of the insertion hole, and a piston rod slidably inserted into the inner side of the sealing sleeve. The bottom end of the piston rod extends into the air storage cylinder and is fixedly connected with a piston head, and the piston head is in sliding fit with the inner side wall of the air storage cylinder. The piston head moves up and down, waste gas is effectively collected and stored, when the piston head moves downwards, air in the gas storage cylinder is exhausted, meanwhile, the waste gas enters the gas storage cylinder from the one-way valve, the design ensures efficient collection of the waste gas, after the switch valve is opened, an inflator is used for inflating the gas storage cylinder, and the waste gas can be effectively collected. Waste gas can be compressed in the gas storage cylinder, effective compression and storage of the waste gas are achieved, and follow-up transportation and treatment are facilitated.

The invention discloses a joint optimization method for a joint photographic experts group (JPEG) quantization table and a truncation position based on rate distortion, relates to the technical field of JPEG coding optimization, and solves the technical problems that in a JPEG compression scheme, improvement of coding efficiency caused by joint optimization by using the quantization table and the truncation position is neglected, and the optimal quantization table and the optimal truncation position are difficult to find. According to the method, joint optimization is carried out on a quantization table and a truncation position of the JPEG by combining a JPEG compression method from the perspective of evolutionary computation, a strategy that the truncation position is fixed when the quantization table is optimized, a strategy that the quantization table is unchanged when the truncation position is optimized, an estimation method for designing fitness evaluation and a method based on rate-distortion optimization are respectively carried out, and the quantization table and the truncation position are optimized. The optimal quantization table and the truncation position can be effectively found, and the corresponding truncation position is optimized while multiple groups of quantization tables are generated in the optimization process; through joint optimization of the quantization table and the truncation position, image data can be compressed more effectively, redundant information is reduced, and therefore coding efficiency is improved.

The invention provides a large model cue word compression method based on dynamic budget allocation. The method aims at optimizing the cue word length of a large language model so as to improve the calculation efficiency of the model and reduce the calculation cost. Through cooperative work of the dynamic budget distributor and the vocabulary selector, the cue words can be compressed on the premise of ensuring semantic integrity. Firstly, a dynamic budget distributor dynamically distributes a vocabulary retention budget of each sentence based on the correlation between each sentence and a question; then, the vocabulary selector calculates the retention probability of each vocabulary through the strategy network, and key vocabularies are preferentially retained; and evaluating loyalty and similarity of compression prompts through a reward function module, and guiding a strategy gradient to update network parameters. According to the method, the performance of the large language model is remarkably improved, and the method is suitable for prompt optimization in scenes such as multi-round dialogues and long text analysis.

The utility model relates to the technical field of stainless steel product production, in particular to a cutting device for stainless steel product production, which comprises a plurality of first conveying rollers and a plurality of second conveying rollers, two material pressing tables are arranged on the workbench between the adjacent first conveying rollers and second conveying rollers at intervals; the cutting assembly is arranged below the workbench and can stretch into the position between the two material pressing tables to cut off the materials. The two limiting assemblies are arranged on the two sides of the workbench correspondingly. The material pressing assembly is arranged above the workbench and comprises two pressing plates horizontally arranged above the two material pressing tables respectively, two first racks are vertically arranged on the upper surfaces of the two material pressing tables respectively, and the upper ends of the two first racks are connected with a top plate above the workbench in a sliding mode and connected with the lifting component after penetrating through the top plate. The cutting device has the advantages that continuous cutting operation is facilitated, materials are stably positioned during cutting, and the cutting effect is good.

The invention discloses a food packaging film sealing machine, and relates to the technical field of film sealing machines, the food packaging film sealing machine comprises a film sealing rack, the top of the film sealing rack is provided with a belt conveying rack, the top of the inner side of the film sealing rack is fixedly connected with two sets of suspension brackets, and the inner side of each set of suspension bracket is rotationally connected with a fixed roller; and a packaging film is wound between the two fixed rollers. According to the food packaging box, through cooperation of parts such as the gas outlet channel, inert gas is input into the gas outlet channel from the hose and then blown out of the gas outlet of the gas outlet channel, water stains and other stains on the surface of the food packaging box are removed, and the clean surface is beneficial to tight attachment between the packaging film and the food packaging box; the purging of the inert gas ensures that no impurity interference exists in the packaging area, so that the heat sealing head can uniformly and effectively heat and press the packaging film, and the sealing performance and firmness of packaging are improved.

The invention discloses an LLMs combined compression method based on joint optimization, an electronic device and a storage medium, a quantization method and a rarefaction method are combined, errors caused by quantization and rarefaction are minimized at the same time, the method comprises the following steps that 128 fragments are sampled from a WikiText2 training corpus, and each fragment comprises 2048 tags; 2, initializing a diagonal transformation matrix by using a statistical index, carrying out preliminary quantization on a weight by using an equivalent transformation operation, determining a quantization step length and a zero point, and calibrating the quantization step length and the zero point through the data sampled in the step 1; calculating a sparse mask, rearranging weights by using an importance index # imgabs0 #, and preferentially sparsening unimportant weights; carrying out joint optimization; and evaluating the compressed model by using different data sets, and verifying the performance improvement of the model.

The invention belongs to the technical field of point cloud, and relates to a point cloud attribute coding method based on a graph attention network, which comprises the following steps: acquiring point cloud data, and carrying out block division on the point cloud data; performing detail level generation on the point cloud data subjected to block division, predicting point cloud attributes according to a detail level generation sequence, and inputting a neighbor point set with attribute information into a multi-head image attention layer for feature extraction to obtain a prediction residual error of a current point; encoding the prediction residual error of the current point to obtain an attribute code stream; inputting the attribute code stream into a decoding module to obtain a prediction residual error; searching a neighbor point set of unpredicted points in the prediction point set, and inputting the neighbor point set into the attention layer of the multi-head image to predict the current point; combining the predicted value with the decoded predicted residual error to obtain a reconstruction attribute; according to the method, efficient compression and accurate reconstruction of the point cloud data are realized in a layer-by-layer recursive predictive coding mode, and the method can be applied to a three-dimensional scene needing efficient data transmission and storage.

The invention relates to the technical field of computers, and provides a potential space compression method, a potential space decompression method, a potential space compression device, a potential space decompression device, potential space compression equipment and a potential space decompression program product. Compressing the first potential variable through a neural network compression model to obtain a compressed second potential variable, and compressing the first potential variable through a neural network down-sampling model to obtain a down-sampled first potential variable; the arithmetic coding model is used for carrying out lossless Gaussian arithmetic coding on the down-sampled first potential variable to obtain a coded second potential variable; and the compressed second potential variable is transmitted to a decoding end, so that the decoding end decodes the second potential variable to obtain a decoded target image, the structural information of the data can be effectively utilized, efficient lossless compression can be realized, the first potential variable can be compressed more efficiently, the compression ratio of the image data is further improved, and the compression efficiency of the image data is improved. And image data compression cost is reduced.

The invention discloses an index sequence generation method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring first gene expression data and at least one codebook entry, encoding the first gene expression data, extracting local information in the first gene expression data, and mapping the local information into a potential space to obtain potential representation of the local information; performing quantization processing on the potential representation based on the at least one codebook entry to generate a first codebook index matrix and at least one quantization vector; decoding and reconstructing the first codebook index matrix, outputting second gene expression data, performing adversarial training on a decoder and at least one quantization vector, and outputting optimization parameters; and performing autoregression modeling processing on at least one new codebook entry to obtain a second codebook index matrix, and processing the second codebook index matrix by using the optimized decoder to obtain a data sample. According to the method, the calculation burden of high-dimensional data modeling is reduced, and the data capturing capability of the model is improved.

The invention discloses a quantitative reasoning method and system of an AI model and computing equipment. When the AI model is subjected to quantitative reasoning, the reasoning system distributes the weight parameters of the column where the outliers are located in the weight matrix to one or more calculation units in a centralized mode, and distributes the weight parameters which do not contain the column where the outliers are located in the weight matrix to other calculation units, so that batch processing of the outliers is facilitated, and the processing efficiency is improved. And the expenditure caused by outlier processing when other calculation units carry out quantitative reasoning is reduced. The AI model separates the weight parameter of the column where the outlier is located from the weight parameter which does not contain the column where the outlier is located, different data formats are adopted for processing, and high-efficiency compression is achieved on the premise that precision is guaranteed.

The invention provides a roll bending combined die for a carrier rocket special-shaped frame with multiple sections and a using method, the roll bending combined die comprises a middle front roller, a middle rear roller, a side roller, a side pressing wheel and a plurality of roller cushion blocks, the middle front roller is installed on the front side of the middle of the special-shaped frame, and the middle rear roller is installed below the rear side of the middle of the special-shaped frame. Side rollers and side pressing wheels are installed on the two sides of the special-shaped frame respectively, and roller cushion blocks are installed below the middle front roller and the middle rear roller. The method has the beneficial effects that a split combined die structure is adopted for roll bending forming of the profile frame, the number of dies is reduced, the production cost is reduced, and workers can conveniently transport and install the profile frame; the split combined die structure can select whether a cover plate needs to be installed or not according to profile characteristics, friction of a thin area at the top end of a part is reduced, and part forming quality is improved.

The invention provides a data compression method and device for an oblique photography model, equipment and a storage medium, and relates to the technical field of geographic data processing, and the method comprises the steps: obtaining data of a to-be-compressed oblique photography model, analyzing the data to obtain a vertex list and an index list of model data, and creating a linked list based on the vertex list and the index list, first vertexes included in the vertex list and indexes corresponding to the first vertexes are stored in the linked list; constructing a detection window based on the resolution of the to-be-compressed oblique photography model, wherein the size information of the detection window is determined based on the resolution of the to-be-compressed oblique photography model; and traversing each first vertex of the to-be-compressed oblique photography model through a vertex merging detection window to determine a first vertex needing to be merged, and compressing data of the oblique photography model based on an index corresponding to the first vertex needing to be merged. According to the invention, effective compression of the data is realized, and the transmission and storage efficiency of the to-be-compressed oblique photography model data is improved.

The utility model discloses a floating stile sash structure with a heat insulation function, which comprises a window sash part and a floating stile sash part, the window sash part comprises an upper window sash plate and a lower window sash plate, and the upper window sash plate is connected with the lower window sash plate through a first heat insulation plate; the floating stile fan part comprises an upper floating stile plate and a lower floating stile plate, and the upper floating stile plate and the lower floating stile plate are connected through a second heat insulation plate. According to the floating stile sash structure with the heat insulation function, the window sash part and the floating stile sash part are respectively formed by assembling various parts, so that the machining difficulty of metal profiles is reduced, and the production efficiency is improved. By arranging the heat insulation plate, the heat insulation effect of the door and window is improved. By arranging the wire pressing plates on the two sides, installation is easy and fast, the service life is long, the rubber strip can be effectively pressed, meanwhile, rich shapes are provided, and the attractiveness of the door and window floating stile sash is improved.

The double-peak countercurrent double-cavity tuyere comprises a tuyere small sleeve body, a first front cavity water chamber is formed in the right end of the tuyere small sleeve body, and a second front cavity water chamber is formed in the position, located on the outer side of the first front cavity water chamber, of the right end of the tuyere small sleeve body. Through the arrangement of the flow guide partition plate, the interior of the flow guide box can be divided into two groups of different water chambers, the time required for cooling water to flow through the interior of the flow guide box is effectively prolonged, meanwhile, through the arrangement of the flow guide partition plate, the multiple groups of communicated water chambers are formed between the flow guide box and the tuyere small sleeve body due to the spiral shape of the flow guide partition plate, and the cooling effect is improved. And cooling water flowing through the first flow guide holes in the flow guide box can be effectively and spirally guided, the time for cooling the tuyere small sleeve body in the flowing process of the cooling water is further prolonged, and therefore the cooling effect on the tuyere small sleeve body is greatly improved.