96 results about "Linear complexity" patented technology

The invention discloses a physical information neural network-driven ocean island typhoon scene generation method and system. The method comprises the steps of collecting multi-source heterogeneous meteorological data and performing space-time alignment preprocessing; constructing a coarse-scale space-time probability prediction model, capturing space correlation of meteorological elements by using a graph topology learning network, efficiently processing long-time-sequence dependence of typhoon evolution by integrating a state space model with linear complexity, and generating a probabilistic typhoon scene with coarse resolution through a multivariable joint distribution probability model; further constructing a physical downscaling model, taking a coarse-scale prediction result as condition input, and performing physical consistency downscaling on a coarse-scale scene by embedding an atmospheric fluid mechanics equation in a loss function as a physical hard constraint; and finally, outputting a high-resolution typhoon scene with probability reliability and physical authenticity.

The invention relates to the technical field of intelligent agriculture and computer vision, in particular to a small-sample high-density chicken counting framework based on a deep learning Mamba structure, which comprises a feature extraction network, a support-query enhancement module and a decoder. According to the method, a multi-scale feature extraction network based on a residual block (ResNet Block) is introduced, so that local detail information is effectively reserved; and then, a support-query enhancement module is constructed by introducing a Mamba structure, and context interaction between support features and query features is effectively enhanced by utilizing the long sequence modeling capability of linear complexity of the support-query enhancement module, so that the problems of individual overlapping and boundary fuzziness in a high-density chicken flock scene are solved. Experimental results on a PoultryCount real breeding data set show that the mean absolute error (MAE) and the root-mean-square error (RMSE) of the method are reduced compared with those of an existing method, and the chicken counting precision and generalization ability under the condition of a small number of labeled samples are remarkably improved.

The invention relates to the technical field of quantum computing, and provides a linear complexity quantum state preparation method based on tensor decomposition and a quantum circuit construction system.The high-dimensional tensor is decomposed into a series of low-rank core tensors through continuous singular value decomposition, each core tensor in a core tensor sequence is expanded into a unitary matrix, and the unitary matrix is used as a quantum circuit; the unitary matrix sequence is mapped to quantum lines coupled using adjacent qubits, and the quantum lines are run to prepare a target quantum state. Based on this, the line generated by the method can approximately or accurately prepare a target quantum state only by coupling adjacent quantum bits, and is perfectly adaptive to quantum chips of linear or grid topologies such as superconducting and semiconductor quantum dots and the like.

The invention relates to the technical field of medical data processing, in particular to a data analysis and processing system based on clinical hearing detection, and aims to solve the problems that dynamic characteristics of auditory nerve response cannot be systematically described, high-dimensional multi-modal feature vectors cannot be constructed, and the data analysis and processing efficiency is low in the prior art. A stable and reliable input basis cannot be provided for intelligent classification, anomaly detection and individualized analysis, and the sensitivity and specificity of hearing state recognition are reduced; the multi-dimensional hearing feature extraction module fuses time domain, frequency domain and non-linear complexity analysis methods, systematically depicts dynamic features of auditory nerve response, constructs a high-dimensional multi-modal feature vector by integrating three-dimensional features, remarkably improves sensitivity and specificity of hearing state recognition, and improves the accuracy of hearing state recognition. The feature representation method provides a stable and reliable input basis for subsequent intelligent classification, anomaly detection and individualized analysis.

The invention discloses a trajectory planning method based on a multi-stage optimization strategy and a mixed diffusion model. The method comprises the following steps: fusing multi-source sensor data and extracting features; screening and fusing intention anchor point tracks; and optimizing the fine-grained trajectory based on the Transform-Mamba mixed diffusion model. According to the method, an intention anchor point track screening mechanism is provided, the anchor point track matched with the scene is selected from the compact vocabulary through the dynamic screening module and fused with the static prior anchor points, the calculation complexity is reduced, the diversity of the initial track and the scene consistency are guaranteed, and the problem that a traditional fixed anchor point set is insufficient in flexibility is solved. According to the method, a mixed diffusion decoder is designed, space environment dependence is efficiently modeled through a cross attention mechanism, a bidirectional Mama module captures time sequence motion dependence with linear complexity, global perception ability and long sequence modeling efficiency are both considered, and safety and dynamics feasibility of generated tracks are ensured.

The invention belongs to the technical field of machine state prediction, and particularly relates to a fault diagnosis method and system based on a coding and decoding attention mechanism enhanced pulse neural network, and the method comprises the steps: collecting a device vibration signal, and segmenting the device vibration signal into time sequence segments; converting each fragment into a two-dimensional time-frequency image by using S transformation to construct a data set; the image is input into a lightweight hybrid network model to extract deep features, the model is composed of a pulse convolution encoder and a pulse efficient additive attention mechanism encoder which are alternately cascaded, and the pulse convolution encoder extracts local features through pulse neurons and depth separable convolution; the global context dependency is modeled by adopting an efficient additive attention mechanism with low linear complexity; and finally, fault classification is completed based on deep features. Through fusion of pulse calculation and an efficient attention mechanism, the feature extraction capability and diagnosis precision are improved while the parameter quantity of the model is remarkably reduced, and the method is particularly suitable for being deployed on edge equipment with limited resources to realize fault diagnosis.

The invention relates to the technical field of computer vision and document understanding, and discloses a table recognition method based on structure perception coding and state space decoding. The method comprises the steps of firstly extracting table image features; then, features are enhanced through structure perception coding, two-dimensional position coding and row and column embedding are superposed to form grid prior, and serialized visual representation is constructed to improve modeling of a global structure and a row and column alignment relation and reduce redundant calculation. In the decoding stage, state space decoding with linear complexity is adopted, and a table structure mark is generated through autoregression; when a cell start mark is detected, synchronous prediction of cell positioning and content decoding is triggered, and joint optimization is carried out in training. And finally, carrying out fusion and consistency verification on the structure and content results, and outputting a standardized structure and content alignment grid. The method has high recognition precision, robustness and reasoning efficiency in complex scenes of merging cells, invisible boundaries, irregular layout, super-long sequences and the like.

The invention relates to a medical image segmentation system and method based on context and frequency guidance and a storage medium, and the system employs an encoder-decoder architecture, and integrates three core modules: a visual state space module which captures global context information with linear complexity based on a selective state space model; the frequency-guided representation module is used for explicitly separating and enhancing structure and boundary characteristics through frequency domain transformation and complex weight modulation; and the multi-scale adaptive context aggregation module is used for integrating multi-scale semantics through parallel multi-branch convolution and dual Top-K sparse attention and focusing a key region. According to the method, the technical problems that global modeling and boundary details are difficult to consider, the calculation complexity is high and the adaptability to multi-scale targets is poor in the prior art are solved, the segmentation precision and the boundary description accuracy are remarkably improved in various medical image segmentation tasks such as heart MRI, polyp, skin lesion and pathological sections, and the medical image segmentation efficiency is improved. And meanwhile, the calculation efficiency is ensured.

The invention discloses a double-flow fault diagnosis method based on state space modeling and dynamic edge context awareness, the method constructs a time domain state flow and a time frequency edge flow, and the specific implementation process is as follows: the time domain state flow is subjected to state space modeling under a Transform framework to form a MambaT encoder, long and short term dependence is efficiently captured with linear complexity, and the time domain state flow is subjected to dynamic edge context awareness; time domain sensitive features are accurately represented; the time-frequency edge stream converts vibration data into a two-dimensional time-frequency image through wavelet transformation, a differentiable edge sensitive mask is designed to adaptively separate time-frequency image impact characteristics and background noise, boundary content perception adaptive filling is designed to maintain image boundary topology integrity, and the two parts cooperate to form dynamic edge context perception convolution, so that the image boundary topology integrity is improved. Time-frequency edge features are effectively extracted; and a channel enhancement and channel gating module is designed to establish a dynamic channel interaction gating mechanism, so that time domain-time frequency domain double-current cascade characterization is realized, and the fault discrimination capability is remarkably improved. Experiments prove that the method has higher fault diagnosis performance.

The invention belongs to the technical field of medical image segmentation, and relates to a dual-path fusion neural network for prostate precise segmentation and a segmentation method, the neural network constructs a dual-path decoupling encoder architecture based on an nnU-Net framework, captures fine anatomical structure and local texture information through a context sensing residual encoder of a local path, and obtains a dual-path fusion neural network for prostate precise segmentation. A long-range dependency relationship is modeled with linear complexity through a visual state space module of a global path, a double-flow alignment gating module is designed to realize self-adaptive alignment and fusion of cross-path features, and model training is optimized in combination with a mixed loss function and a depth supervision strategy; according to the method, the limitation of an existing segmentation model in the aspect of local structure and global semantic integration is solved, and the segmentation precision, the boundary goodness of fit and the generalization performance of the focus and gland region in the prostate MRI image are improved.

The invention discloses a question and answer method and system based on large model causal graph discovery and causal graph enhanced reasoning, and belongs to the technical field of natural language processing and artificial intelligence. The method comprises the following steps: firstly, carrying out hierarchical expansion from an initial root node by adopting a breadth-first search strategy, and efficiently constructing a reliable causal graph conforming to directed acyclic graph constraints with linear complexity in combination with real-time loop detection; in the question and answer stage, key entities in user questions are analyzed, and causal paths connecting the key entities are retrieved in a causal graph; and taking the retrieved structured causal path as a constraint condition to be injected into a decoding process of a large-scale language model, and generating a natural language answer which is strict in logic and can trace reasoning steps. According to the method, the resource consumption of large-scale causal discovery is remarkably reduced, the accuracy and interpretability of the answers of the questions and answers are effectively improved, and technical support is provided for medical treatment, finance and other scenes needing high-reliability reasoning.

The invention provides an efficient single-stage super-resolution method and a remote sensing image super-resolution system for RGB remote sensing images, and the method integrates Fourier position coding guided geometric node modeling, hypergraph high-order relation modeling and a Mama long-range dependence modeling mechanism based on a state space model. And consistent modeling of a multi-scale structure is realized while the linear calculation complexity is kept. In the method, FOPE is introduced as explicit geometric priori to guide adaptive screening of key nodes on global, regional and local scales; then, an FOPE enhanced hypergraph network is constructed, and high-order space association among multiple nodes is modeled through hyperedge; according to the method, a two-stage Mama aggregation module is further designed, and intra-node feature consistency modeling and inter-node long-range spatial dependency modeling are completed with linear complexity; and finally, a high-resolution RGB remote sensing image is generated through a lightweight direct up-sampling module.

The invention relates to an RGB-T target tracking method and system based on spatio-temporal state evolution, and the method comprises the steps: constructing a double-branch RGB-T target tracking model, enabling RGB and TIR branches to share the weight of a ViT encoder, employing an iterative processing frame, and transmitting spatio-temporal context information between frames through an updatable context memory Tokens. In each iteration, after RGB and TIR modal features are extracted respectively, cross-modal time sequence context modeling is carried out through a modal perception time sequence Mamba module, and the module realizes long-time target perception representation learning through a cross-modal coupling state transition mechanism and a prompt guide strategy; and carrying out intra-modal and inter-modal feature fusion in a spatial dimension through a cross-modal Mama aggregation module, and finally outputting a target position through a prediction head. According to the method, lasting cross-modal state evolution can be realized with linear complexity, the spatial-temporal characteristics of visible light and infrared light are effectively fused, and the tracking robustness and accuracy are improved in complex scenes such as rapid target movement, shielding or modal degradation.

The present application relates to the field of artificial intelligence. Provided in the present application are a method and system for implementing a self-attention mechanism having a linear complexity, and a device and a medium. The method comprises: performing positional encoding on sequence data to be processed, and performing feature mapping to apply the positional encoding to a query matrix and a key matrix; on the basis of the query matrix and the key matrix, performing calculation to obtain a low-rank query matrix, a low-rank key matrix and a low-rank generalized inverse matrix; augmenting a value matrix to obtain an augmented matrix, and on the basis of the augmented matrix, the low-rank key matrix, the low-rank generalized inverse matrix and the low-rank query matrix, obtaining a sequence self-attention matrix; and transposing the sequence self-attention matrix, using the transposed matrix as sequence data to repeat the foregoing steps, using the newly obtained sequence self-attention matrix as a feature self-attention matrix, and on the basis of the sequence self-attention matrix and the feature self-attention matrix, obtaining a self-attention matrix having a linear complexity. The present application implements linear self-attention, better maintains the model performance and has relatively good expansibility.

The invention discloses a traffic flow prediction method and system based on CT-SSM and two-way dynamic graph convolution, and the method comprises the steps: constructing an urban road network topological graph, and constructing a multi-scale periodic historical traffic flow sequence; inputting the multi-cycle fusion feature tensor into a time sequence feature extraction module based on CT-SSM, and based on a selective parameter driven by data, performing state recursion after discretization of a zero-order retainer, and extracting time sequence features; based on a space-time attention mechanism and static mask constraints, generating a global implicit semantic graph and a local physical evolution graph in parallel; performing spatial depth feature aggregation through double-path collaborative dynamic graph convolution; and the dynamic gating unit carries out weighted fusion and then sends the fused data to a prediction result output module to give a prediction result, and model training is completed based on back propagation. According to the method, long-range time sequence dependence is extracted through linear complexity, time-space dependence is modeled through time-space attention depth, topology solidification is broken through construction of a two-way dynamic graph, and the precision and robustness of traffic flow prediction are remarkably improved.

The invention discloses a power system moving target defense method based on power flow betweenness and power flow disturbance, and relates to the field of power systems. Calculating to obtain a power flow betweenness and a power flow disturbance index of the normalized power system line; constructing a comprehensive index based on the power flow betweenness and the power flow disturbance index; and determining a to-be-deployed line set according to a preset device deployment number and the comprehensive index. And changing the line impedance value corresponding to the line set to obtain a target measurement matrix, calculating a measurement value residual error after the line impedance value is changed according to the original measurement matrix and the target measurement matrix, and determining an attack success rate based on a size relationship between the measurement value residual error and a detection threshold value. And deploying power system moving target defense according to a result of the attack success rate. According to the invention, the detection effectiveness and the MTD concealment are balanced. A traditional global combination search problem is converted into a linear complexity sorting problem, rapid solving is achieved, calculation efficiency is improved, and rapid generation of a hidden MTD strategy is achieved.

The application discloses a kind of based on double-domain calibration image segmentation method and system, method uses double-flow feature extraction network, first branch extracts local space feature to obtain first local feature;Second branch is converted to frequency domain after extracting second basic feature, and noise filtering is carried out to obtain frequency domain calibration feature by spectrum texture calibration module, spatial boundary calibration module is extracted in spatial domain Structure boundary response obtains spatial domain calibration feature, then linear attention mechanism is aggregated with second basic feature to obtain long-range global feature;First local feature and long-range global feature are adaptively fused to output segmentation result.The application is filtered out by double-domain collaborative calibration of frequency domain and spatial domain, and weak boundary response is enhanced, linear complexity global context modeling is realized by combining linear attention, and the robustness and accuracy of image segmentation in strong noise and weak boundary scene are significantly improved.

The application provides a method and device for graph data processing, and relates to the technical field of artificial intelligence, and the method comprises: a method for graph data processing, which comprises the following steps: acquiring graph data, extracting the basic structure features of the graph data through a static topology path, adaptively screening the key neighbor features in the graph data based on an attention mechanism and in combination with the basic structure features through a dynamic path, establishing long-distance semantic association between nodes through topology-aware coding based on the basic structure features and the key neighbor features through a linear graph transformer with O ( n ) of linear complexity, and establishing long-distance semantic association between nodes through topology-aware coding; performing frequency domain projection on the node features to generate frequency domain features, explicitly extracting and enhancing the periodic discriminant components in the frequency domain features through a Fourier analysis network to generate graph neural network data. Thus, the graph representation learning capability of the graph neural network is effectively improved.

The invention discloses an instance segmentation method in a low-contrast environment and a readable storage medium, and belongs to the technical field of image processing, the instance segmentation method in the low-contrast environment introduces an adaptive state space module, performs long-range dependence modeling based on a down-sampled input image to obtain a state space feature, and obtains a state space feature; and performing multi-scale edge feature extraction based on the state space features by adopting a multi-scale adaptive edge enhancement module to obtain multi-scale edge enhancement features, screening a final prediction frame of the target object through an instance segmentation network based on the multi-scale edge enhancement features, generating a final mask of the target object, and obtaining the final mask of the target object. And the target object is subjected to instance segmentation according to the final mask, so that sequence modeling of linear complexity is realized, excellent long-range dependence modeling capability is obtained while efficient calculation is kept, and the edge features of the object to be subjected to instance segmentation can be well adaptively enhanced in a low-contrast situation.

This invention relates to a dual-branch remote sensing image semantic segmentation method and system based on Mamba and adaptive convolution, belonging to the fields of remote sensing image semantic segmentation and artificial intelligence technology. This method aims to solve the problems of existing technologies, such as difficulty in coordinating global dependencies and local texture information, low segmentation accuracy at multiple scales, poor multi-band adaptability, and low training efficiency. The invention includes using a multispectral adaptive processor for spectral grouping and attention enhancement, inputting the data into a dual-path backbone network after downsampling through a convolutional backbone. The global path uses a Mamba module to achieve linear complexity global modeling, while the local path uses adaptive convolution to dynamically adjust weights to capture details. A multi-head cross-attention fusion module then enables bidirectional interaction of multi-scale features, and a feature pyramid network performs cross-scale optimization and resolution restoration. This invention significantly improves segmentation accuracy, reduces computational burden, and adapts to the real-time application requirements of high-resolution remote sensing images.

The invention discloses a multi-category industrial anomaly detection method based on a state space model. The method comprises the following steps: 1, collecting and preprocessing image data of multi-category industrial products; 2, constructing an encoder network to encode the image to obtain a multi-scale semantic feature map; 3, constructing a multi-scale feature fusion network to reduce feature redundancy and strengthen network characterization capability; 4, constructing a decoder network to reconstruct multi-scale features; and 5, detecting an abnormal region according to the reconstruction error. The multi-class industrial anomaly detection method based on the state space model is different from a single-class anomaly detection method, and expresses accurate anomaly classification and positioning effects on multiple classes of industrial products at the same time; different from a convolutional neural network and a self-attention mechanism, the method utilizes a state space model to realize global relation modeling and improve the detection efficiency according to the characteristics of keeping linear complexity along with the input size.

The invention belongs to the technical field of ship electric power, and discloses a method and a system for quickly deducing a damaged state of a ship electric power system, and the method comprises the steps: carrying out the modeling of a tree-shaped Bayesian network; and quickly calculating the joint probability of the outage combination. According to the method, a rapid inference algorithm of linear complexity is provided by combining a tree operation structure of a power system, and the exponential complexity O (2n) of a traditional method is reduced to the linear polynomial complexity O (n). The method is realized by calculating the joint probability of the outage combination in a recursion mode: matrix multiplication operation is only required to be performed on each node once, and the total operation times and the node number n are in a linear relation. For an 86-node ship power system, the deduced time consumption is reduced from several hours of a traditional method to less than 0.75 second, and the efficiency is improved by more than 10000 times.

The RWKV-based novel translation model inference method and system includes the following steps: 1) Collecting novel translations and extracting parallel corpora, and using dynamic MicroBatch concatenation technology for sequence compression; 2) Introducing a lightweight group query attention mechanism on the basis of the RWKV architecture, directly obtaining KV information from the Embedding layer to build the model; 3) Employing a sublinear complexity hybrid parallel training mode, combined with a global scalar scaling FP16 mixed precision strategy for training; 4) Applying a hierarchical distributed heterogeneous architecture, offloading the optimizer to low-performance devices and performing gradient compression transmission; 5) Outputting the translation using a joint decoder and dynamic batch inference technology. This invention, through the above method and system, effectively reduces the computational complexity and memory usage in the long text translation process, improves model training efficiency and inference throughput, and significantly improves the translation efficiency and contextual coherence of ultra-long texts.

The invention discloses an aero-engine blade ablation detection method based on a double-path state space model, and belongs to the technical field of computer vision and aero-engine intelligent detection. The core of the method is that a double-path neural network model is adopted, a plurality of visual state space VSS modules are integrated on a left path, and long-range dependence modeling is achieved with linear complexity; and the right path extracts local features based on the ResNet50 network. Dual-path output features are integrated through a multi-scale feature fusion module, and finally semantic segmentation is completed through a PSPNet solution terminal. According to the method, the problems of high calculation complexity and insufficient boundary segmentation precision in the prior art are effectively solved, and the segmentation precision and robustness of the ablation region, especially weak textures and irregular boundaries, are remarkably improved while the calculation overhead is reduced.

The invention discloses a large-scale visual positioning optimization method based on an OCP theory, and relates to the technical field of deep learning. The method comprises the following steps: defining a target function of a visual positioning model; initializing model parameters, momentum vectors and related hyper-parameters; performing iterative optimization training, calculating a small-batch stochastic gradient in each iteration, performing exponential moving average and deviation correction on the gradient by using a diagonal element of an approximate Hessian matrix of element-by-element square of the gradient, performing weight attenuation, and finally calculating a parameter update quantity and updating a model by using an optimization method based on an OCP theory; and outputting the model with the optimal performance on the verification set after iteration is finished. According to the method, the OCP theory and approximate second-order information are combined, a new large-scale visual positioning method is provided, the convergence speed, stability and final test precision of visual positioning model training are effectively improved on the premise that linear complexity is kept, and the method is particularly suitable for large-scale non-convex optimization scenes.

The invention discloses a sequence processing and optimization method and system based on adaptive sparse gating, electronic equipment and a storage medium, and relates to the technical field of artificial intelligence and deep learning. The method comprises the following steps: executing inertial evolution, performing permanent operation with linear complexity by using an inertial processing unit, and updating a hidden state; entropy judgment is carried out, information entropy reflecting prediction uncertainty is calculated, and a gating coefficient is generated; executing on-demand activation, physically bypassing a geometric correction unit at a common time according to a gating coefficient, activating the unit at a critical time, and loading a historical key value pair to perform global self-attention calculation; and finally, manifold fusion is carried out based on the gating coefficient to generate an output state. According to the method, through an adaptive sparse calculation mechanism, the reasoning cost is greatly reduced, and meanwhile, the high precision and anti-forgetting ability of the model in a complex task are effectively guaranteed.

The application discloses a kind of physical information neural network driven ocean island typhoon scene generation method and system, comprising: collecting multi-source heterogeneous meteorological data and carrying out space-time alignment preprocessing;Build coarse scale space-time probability prediction model, utilize graph topology learning network to capture the spatial correlation of meteorological elements, integrate the state space model of linear complexity to efficiently process the long-time sequence dependence of typhoon evolution, and generate coarse resolution probabilistic typhoon scene through multivariate joint distribution probability model;Further build physical downscaling model, input coarse scale prediction result as condition, embed atmospheric fluid mechanics equation as physical hard constraint in loss function, and physically consistent downscaling is carried out to coarse scale scene;Finally output high-resolution typhoon scene with probability reliability and physical authenticity.

The application discloses a three-dimensional point cloud change detection method based on state interaction fusion Mamba network, which comprises the following steps: carrying out block and Z-order serialization on double-phase point clouds, and inputting the double-phase point clouds into a Fusion Mamba module; generating selective parameters of a state space model of another phase by using features of one phase, modulating a state evolution process through a bidirectional symmetry mechanism, realizing deep cross-phase interaction fusion, returning features and carrying out differential classification to obtain a change result; the application utilizes the linear complexity characteristic of the Mamba framework, replaces traditional feature splicing through state interaction at the parameter level, effectively models long-range dependence, and completes three-dimensional point cloud change detection; the application significantly reduces the calculation complexity to O (N), solves the problem of high memory occupation of the Transformer method, and improves the discrimination of double-phase feature fusion through deep state modulation, thereby realizing high-precision and low-resource-consumption change detection in a large-scale urban scene.

The invention discloses an image compression method based on a state space, and belongs to the technical field of image compression. The method comprises the following steps: constructing a hybrid model comprising a nonlinear analysis transformation network, a nonlinear synthesis transformation network, a hyper-priori network and an entropy model network, wherein the model fuses a state space model and a CNN to maintain linear complexity; inputting a to-be-coded image into the nonlinear analysis transformation network and converting the to-be-coded image into a hidden state feature; processing the hidden state feature by using a hyper-prior network to obtain quantized, coded and decoded up-sampling hyper-prior potential representation; inputting the representation into an entropy coding network, predicting Gaussian distribution of each channel of the hidden state feature, and performing arithmetic coding and decoding; and inputting the decoded features into a nonlinear synthesis transformation network to finally obtain a reconstructed image. According to the method, the compression performance is effectively improved while the model complexity is reduced.

The invention relates to the technical field of biomedical signal processing and artificial intelligence, and discloses a biomedical signal denoising method and system based on a double-branch state space model and phase perception. According to the method, a de-noising model based on a SpectroMama-UNet is constructed, a U-shaped framework is adopted, each layer is integrated with a SpectroMama module, and modeling is carried out on non-stationary signals through parallel time domain branches and phase sensing spectrum branches. The time domain branch captures signal time dynamic characteristics by using a bidirectional state space model; a phase sensing spectrum branch adopts a real and imaginary part splicing strategy, a real part and an imaginary part of a complex frequency spectrum are spliced in a channel dimension and jointly learned, phase information is explicitly reserved, and the problem of waveform drift of a traditional frequency domain method is solved. According to the method, the linear complexity of the state space model is used for replacing the secondary complexity of a traditional Transform, accurate removal of artifacts such as electro-oculogram and myoelectricity and high-fidelity reconstruction of signal waveforms are achieved, and the method is suitable for real-time signal processing of portable medical equipment.