Auxiliary Set Research Articles

Cross-Channel Specific-Mutual Feature Transfer Learning for Motor Imagery EEG Signals Decoding.

In recent years, with the rapid development of deep learning, various deep learning frameworks have been widely used in brain-computer interface (BCI) research for decoding motor imagery (MI) electroencephalogram (EEG) signals to understand brain activity accurately. The electrodes, however, record the mixed activities of neurons. If different features are directly embedded in the same feature space, the specific and mutual features of different neuron regions are not considered, which will reduce the expression ability of the feature itself. We propose a cross-channel specific-mutual feature transfer learning (CCSM-FT) network model to solve this problem. The multibranch network extracts the specific and mutual features of brain's multiregion signals. Effective training tricks are used to maximize the distinction between the two kinds of features. Suitable training tricks can also improve the effectiveness of the algorithm compared with novel models. Finally, we transfer two kinds of features to explore the potential of mutual and specific features to enhance the expressive power of the feature and use the auxiliary set to improve identification performance. The experimental results show that the network has a better classification effect in the BCI Competition IV-2a and the HGD datasets.

Hypergraph contrastive learning for recommendation with side information

PurposeThis paper addresses the limitations of current graph neural network-based recommendation systems, which often neglect the integration of side information and the modeling of complex high-order interactions among nodes. The research motivation stems from the need to enhance recommendation performance by effectively utilizing all available data. We propose a novel method called MSHCN, which leverages hypergraph neural networks to integrate side information and model complex interactions, thereby improving user and item representations.Design/methodology/approachThe MSHCN method employs a hypergraph structure to incorporate various types of side information, including social relationships among users and item attributes, which are essential for enriching user and item representations. The k-means clustering algorithm is utilized to create item-associated hypergraphs, while sentiment analysis on user reviews refines the modeling of user interests. Additionally, hypergraphs are constructed for user-user and item-item interactions based on interaction similarity. MSHCN also incorporates contrastive learning as an auxiliary task to enhance the representation learning process.FindingsExtensive experiments demonstrate that MSHCN significantly outperforms existing recommendation models, particularly in its ability to capture and utilize side information and high-order interactions. This results in superior user and item representations and improved recommendation performance.Originality/valueThe novelty of MSHCN lies in its use of a hypergraph structure to integrate diverse side information and model intricate high-order interactions. The incorporation of contrastive learning as an auxiliary task sets it apart from other hypergraph-based models, providing a significant enhancement in recommendation accuracy.

Structure of Multi-State Correlation in Electronic Systems.

Beyond the Hohenberg-Kohn density functional theory for the ground state, it has been established that the Hamiltonian matrix for a finite number (N) of lowest eigenstates is a matrix density functional. Its fundamental variable─the matrix density D(r)─can be represented by, or mapped to, a set of auxiliary, multiconfigurational wave functions expressed as a linear combination of no more than N2 determinant configurations. The latter defines a minimal active space (MAS), which naturally leads to the introduction of the correlation matrix functional, responsible for the electronic correlation effects outside the MAS. In this study, we report a set of rigorous conditions in the Hamiltonian matrix functional, derived by enforcing the symmetry of a Hilbert subspace, namely the subspace invariance property. We further establish a fundamental theorem on the correlation matrix functional. That is, given the correlation functional for a single state in the N-dimensional subspace, all elements of the correlation matrix functional for the entire subspace are uniquely determined. These findings reveal the intricate structure of electronic correlation within the Hilbert subspace of lowest eigenstates and suggest a promising direction for efficient simulation of excited states.

Exchange-correlation kernel for perturbation dependent auxiliary functions in auxiliary density perturbation theory

ContextAnalytic exchange-correlation kernel formulations are of the outermost importance for density functional theory (DFT) perturbation calculations. In this paper, the working equation for the exchange-correlation kernel of the generalized gradient approximation (GGA) for perturbation dependent auxiliary functions is derived and discussed in the framework of auxiliary density functional theory (ADFT). The presented new formulation is extended to the unrestricted approach, too. A comprehensive discussion of the implementation of the GGA ADFT kernel, using either the native exchange-correlation functional implementations in deMon2k or the ones from the LibXC library, is given. Calculations with analytic exchange-correlation kernels are compared to their finite difference counterparts. The obtained results are in quantitative agreement. Nevertheless, analytic GGA ADFT kernel implementations show substantial improvement in the computational performance. Similar results are reported for analytic second derivatives of effective core potential (ECP) and model core potential (MCP) matrix elements when compared to their finite difference counterparts in molecular frequency analyses.MethodAll calculations are performed in the framework of ADFT as implemented in deMon2k. In the ADFT analytic frequency calculations, auxiliary density perturbation theory was used. The underlying two-center exchange-correlation kernel matrix elements are calculated by numerical integration either with analytic or finite difference kernel expressions. Validation calculations are performed with the VWN and PBE functionals employing DFT-optimized DZVP basis sets in conjunction with automatically generated GEN-A2 auxiliary density function sets. In the (Pt3Cu)n cluster benchmark calculations, the RPBE functional was used. For Pt atoms, the quasi-relativistic LANL2DZ effective core potential with the corresponding valence basis set was employed, whereas for Cu atoms, the all-electron DFT-optimized TZVP basis was applied. The auxiliary density was expanded by the automatically generated GEN-A2* auxiliary function set. We run all benchmark calculations in parallel on 24 cores.

Оценка нестабильности последовательных впрыскиваний топливной аппаратуры дизелей

The issues of uneven cyclic feed in sequential injections by diesel fuel equipment are considered, which is directly related to solving the problem of ensuring the quality of operation of engines of various types and purposes when operating at low loads and rotational speeds. The special importance of solving these problems from the standpoint of ensuring fuel efficiency, environmental pollution, solving issues of engine reliability and improving their operational properties was noted. The features of the formation of residual fuel pressure in the high-pressure pipeline are revealed. The periodic fluctuations of the cyclic feed and injection parameters from cycle to cycle are estimated. A study of possible options for the operation of fuel equipment in low fuel supply modes has been carried out. Based on the analysis, a criterion for evaluating the instability of successive injection cycles for diesel fuel equipment during its operation in various modes has been introduced. By studying numerical examples, the conditions of unstable fuel supply processes in successive injections of various characteristic modes are determined. The criterion determines the quality of the fuel equipment of marine diesel engines and diesel generators in terms of ensuring a minimum stable idle speed and the possibility of ensuring stable reduced fuel supplies. The variability of fuel supply processes in sequential feeds is the cause of increased exchange processes of electric power flow during parallel operation of diesel alternators and reduces the performance characteristics of diesels for various purposes when operating in partial modes. In the process of carrying out development work, scientific research and during the refinement of diesels, the instability criterion allows us to assess the quality of the fuel equipment precisely in partial modes. This is of particular importance for diesel engines of fishing vessels operating for a long time at shared loading modes and low idle speeds, as well as for engines of the tugboat fleet and auxiliary diesel generator sets of all types of facilities used.

Supervisory adaptive safety control for uncertain nonlinear systems

This paper investigates the supervisory adaptive safety control problem for uncertain nonlinear systems with unmeasurable states. In order to estimate unknown parameters and unmeasurable states, a supervisory observer framework consisting of a multi-observer bank and a supervisor unit, is proposed based on a dividing rectangles (DIRECT) algorithm. Under a desired accuracy, the parameter estimates converge to near the true parameters within finite-time while the estimation error of the system state converges to a small neighborhood near the origin. Also, an observer robust adaptive control barrier function is constructed to ensure safety for uncertain nonlinear systems by guaranteeing controlled invariance of a time-varying auxiliary set. By adjusting some parameters in the DIRECT algorithm, the time-varying auxiliary set approaches the original safe set, which reduces the conservatism generated by the compression of the safe set. Finally, two examples, including an adaptive cruise control system, are given to demonstrate the effectiveness of the proposed method.

Normal Reorientation for Scene Consistency.

With the remarkable progress of 3D scanning technique, the captured indoor scenes appear increasingly in last decade. Generating orientation-consistent normals for indoor point clouds is a fundamental and important task. The existing orientation rectification methods pay more attention to object-level targets with connected surface. However, it is challenging to compute consistent surface orientation for real scanned indoor point clouds. In this paper, we analyze the causes of this difficulty and propose a new normal reorienting framework for indoor scene consistency, namely NRSC. It first estimates normals for an indoor point cloud and extracts all the connected regions. We then design and construct an abstract orientation bridging tree (OBT) to organize the extracted regions in a hierarchical way. For all node regions, NRSC iteratively implements a set of orientation propagations to generate locally orientation-consistent regions. Moreover, we define an auxiliary viewpoint set for each pairwise parent-child node regions and introduce a voting mechanism to rectify the region orientation of child node according to its parent. After processing all the child node regions along OBT, we finally eliminate the orientation inconsistencies between related regions. Multi-groups of experimental results on both fused indoor scenes and single-view-scenes show that our method generates globally consistent orientation for indoor point clouds.

ON THE RECONSTRUCTION OF PRESTRESS FIELDS IN A HOLLOW CYLINDER

The present research is devoted to the development of the theoretical foundations of nondestructive acoustic method for identifying inhomogeneous prestress fields in a hollow cylinder, depending on the probing loading type. A linearized model of steady oscillations of an elastic body in the presence of an inhomogeneous prestress field of arbitrary nature is considered in the standard and weak formulations. On the basis of this model, we formulate a problem for a cantilever-clamped prestressed hollow cylinder that performs steady axisymmetric vibrations under three types of probing loading. A corresponding weak formulation of the problem in the cylindrical coordinate system is presented, in which six independent components of the prestress tensor are taken into account. At that, a case of prestress fields obtained by applying some initial mechanical external static load is considered. In the presence and absence of prestresses of various types, amplitude-frequency dependences are analyzed, and resonant and natural frequencies are found in a wide frequency range. Numerical calculations were carried out using the FEM on a non-uniform grid; mesh refinement is carried out in the vicinity of the boundary points, where the type of boundary conditions changes. Based on the numerical solution of an auxiliary set of direct problems, seven types of prestress fields are constructed, differing in the types of initial loading, most often encountered in practice. To assess the possibility of implementing the procedure for reconstructing prestresses of each of the considered types, a sensitivity analysis was additionally performed, which showed that for some prestress types there are frequencies and types of probing loading for which the presence of prestress is practically not manifested. The sensitivity analysis performed made it possible to implement the optimal method of probing loading when solving the inverse coefficient problem. The statement of the new inverse problem on the restoration of arbitrary inhomogeneous prestress fields in the considered finite hollow cylinder is formulated. When restoring the prestress of a given structure, the inverse problem is reduced to finding a set of parameters from an ill-conditioned algebraic system, which was studied with the help of the A.N. Tikhonov regularization method. Additional data for solving the inverse problem was obtained on the basis of probing both via a single load and via combined probing modes. It has been found that it is most effective to use a combined loading mode and use a sufficiently wide frequency range when selecting sounding frequencies. The results of computational experiments on the reconstruction of six components of the prestress tensor are presented and analyzed, and recommendations are proposed for choosing the optimal modes of acoustic sounding.

Level set guided region prototype rectification network for retinal vessel segmentation

Retinal vessel segmentation refers to extracting the vessel region with continuous and smooth boundaries from retinal images, which is of great significance in clinical practices. However, due to the weak and blurry edges of targets as well as interference (such as optic cup and disc) in the background, current deep neural network-based methods struggle in extracting features with discriminative semantics while preserving continuous edges. To enforce continuous predictions of weak edges, we propose a level set guided region prototype rectification (LSRPR) framework and a novel level set loss (LS-loss) with learnable and self-guided mechanisms. Specifically, the LSRPR firstly takes features of the last layer from the decoders of a U-Net version as input and rectified the region prototype by an auxiliary self-supervised level set loss, then the pre-trained model is fine-tuned by using supervised level set loss. The LS-loss facilitates the model to generate reliable guidance and enhances the continuous of edges among the decoders of neural network model. The proposed method is simple, yet effective, which can easily be extended to other frameworks. The quantitative and qualitative experimental results on public retinal vessel datasets indicate the effectiveness of the region prototype rectification compared to other SOTA models. Our code is available at Github:https://github.com/tweedlemoon/LSRPR.

FEDAUX: Leveraging Unlabeled Auxiliary Data in Federated Learning.

Federated distillation (FD) is a popular novel algorithmic paradigm for Federated learning (FL), which achieves training performance competitive to prior parameter averaging-based methods, while additionally allowing the clients to train different model architectures, by distilling the client predictions on an unlabeled auxiliary set of data into a student model. In this work, we propose FEDAUX, an extension to FD, which, under the same set of assumptions, drastically improves the performance by deriving maximum utility from the unlabeled auxiliary data. FEDAUX modifies the FD training procedure in two ways: First, unsupervised pre-training on the auxiliary data is performed to find a suitable model initialization for the distributed training. Second, ε, δ)-differentially private certainty scoring is used to weight the ensemble predictions on the auxiliary data according to the certainty of each client model. Experiments on large-scale convolutional neural networks (CNNs) and transformer models demonstrate that our proposed method achieves remarkable performance improvements over state-of-the-art FL methods, without adding appreciable computation, communication, or privacy cost. For instance, when training ResNet8 on non-independent identically distributed (i.i.d.) subsets of CIFAR10, FEDAUX raises the maximum achieved validation accuracy from 30.4% to 78.1%, further closing the gap to centralized training performance. Code is available at https://github.com/fedl-repo/fedaux.

PH-CF: A Phased Hybrid Algorithm for Accelerating Subgraph Matching Based on CPU-FPGA Heterogeneous Platform

Nowadays, more and more data are represented and stored by a graph structure, and subgraph matching is a fundamental problem in a variety of scientific machine learning and industrial applications, such as remote sensing image registration, industrial inspection, etc. Due to the NP-hard problem of subgraph matching, the explosive growth of graph data, the disadvantages of high energy consumption, and the high overhead of CPU and GPU platforms, computing subgraph matching is becoming more and more challenging. To alleviate this problem, we propose a phased hybrid algorithm to accelerate the enumeration task of subgraph matching, called <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">PH-CF</monospace> , based on the CPU-FPGA heterogeneous platform. This approach can make full use of the pipeline and data flow mechanism, low power consumption, and configurable characteristics of FPGA. First, the matching order of query vertices automatically selects GQL or RI methods according to the sparsity of the data (query) graph. Second, a candidate vertex auxiliary data structure set partitioning method is designed to effectively realize the load balance of multiple computing units at the FPGA and CPU host sides. Third, FPGA's pipeline and data flow mechanism is used to accelerate the enumeration phase of subgraph matching. Experimental results on real-world and synthetic datasets show that the performance of the <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">PH-CF</monospace> outperforms the state-of-the-arts. <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">PH-CF</monospace> can obtain the average performance improvement of up to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$16.07\times$</tex-math></inline-formula> , <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$38.61\times$</tex-math></inline-formula> , and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$11.46\times$</tex-math></inline-formula> over CFL, CECI, and DP-iso, respectively. Moreover, our approach has good stability and robustness on various datasets.

Optimized Design of a Multistage Centrifugal Pump Based on Volumetric Loss Reduction by Auxiliary Blades

Throat ring leakage is a major factor deteriorating the performance of multistage centrifugal pumps. This paper focuses on the optimization of multistage centrifugal pumps by incorporating the principle of the Tesla valve and adding an auxiliary set of blades to the impeller body. By changing the direction and magnitude of the leaking fluid’s flow, the leakage volume of the impeller throat ring is reduced. The study results demonstrate that the experimental error in head calculation with numerical simulation at the optimal working condition was 0.65%, verifying the accuracy of the numerical simulation method. The leakage volume of the throat ring decreased by up to approximately 28.99% compared to the original structure, which significantly increased the pump’s head and overall efficiency. Near the optimal operating point, the pump’s head and overall efficiency increased by approximately 8.1% and 8.7%, respectively. The larger the flow rate, the greater the improvement in the pump’s head and total efficiency. Near high-flow operating conditions, the pump’s head and overall efficiency increased by approximately 116.45% and 110.84%, respectively. The auxiliary blade structure introduces a non-contact seal which, compared to traditional seal structures, improves seal life and reduces seal costs. Additionally, the auxiliary blades can shift the optimal operating point of the multistage centrifugal pump towards a higher flow rate, improving the pump’s delivery capability.

An efficient strategy to implement local porosity constraints in the multiscale design of solids with parameterized biomimetic microstructures

In previous works, the authors introduced a multiscale optimization method to maximize the stiffness of elastic solids with biomimetic cancellous microstructures described by a finite set of parameters. Although effective, the procedure is computationally expensive when solving large-scale problems using per-element non-linear constraints to impose local bounds on the solid volume fraction. This work improves the computational performance of the method by exploring two strategies to completely dispense with nonlinear local constraints: to bound the microparameters so the microsctructures are always within the solid fraction of trabecular bone, and to map the microparameters onto an auxiliary set of parameters that are linearly bounded. As a side effect, the design spaces are reduced. Such reductions are assessed in terms of the bulk and shear moduli and elastic symmetries, which are compared to those of natural bone. Performances of the two strategies are assessed by solving a series of benchmark problems and studying the stiffness of a hip prosthesis. The strategy based on the isoparametric mapping achieves the best results, performing up to 2000 times faster while marginally reducing the design space. Thus, the isoparametric mapping approach makes the multiscale design method a suitable tool for solving large-scale problems of practical interest.

Distance‐targeted competitive follower‐attraction containment control for multi‐agent systems with weighted directed graphs

AbstractThis paper studies competitive containment control problems of the leader‐follower multi‐agent systems with weighted directed graphs. Two leaders compete to minimize the ultimate distances from the followers, by directly connecting with some of the followers and allocating weights on these connections. This competition is characterized by a nonzero‐sum game due to the existence of singular strategies. With the help of matrix series and Geršgorin discs, we provide the gradients and the convexity of the payoff functions on the differentiable strategy sets. By constructing auxiliary strategy sets which elaborately exclude the singular strategies and deeply analyzing the Karush‐Kuhn‐Tucker conditions, existence and uniqueness of the Nash equilibrium are rigorously examined, and a sufficient and necessary algebraic condition for this Nash equilibrium is further obtained. For a special case where all followers can be connected with meanwhile a degree condition for the interaction graph is satisfied, we find each leader tends to build connections with all followers and destroy the distribution fashion of multi‐agent systems. All above analysis indicates that some typical phenomena in game theory models such as Pigs' Payoff and Prisoner's Dilemma similarly exist in the multi‐agent containment control systems. Lastly, two Nash equilibrium seeking algorithms are proposed for leaders to select their rational strategies.

Correlation consistent auxiliary basis sets in density fitting Hartree-Fock: The atoms sodium through argon revisited.

We present a series of auxiliary basis sets, for the elements Na to Ar, for use in density-fitted Hartree-Fock calculations with the correlation consistent cc-pV(n + d)Z orbital basis sets. Benchmarking on total molecular energies, reaction energies and the spectroscopic constants of the SO molecule demonstrate that the new sets address the deficiencies of using existing auxiliary sets in combination with these orbital basis sets. We also report auxiliary basis sets for Na and Mg matched to cc-pVnZ, along with recommendations for pairing auxiliary basis sets to the cc-pVnZ-F12 basis sets for Hartree-Fock calculations.

Silent Speech Interface With Vocal Speaker Assistance Based on Convolution-Augmented Transformer

Silent Speech Interface (SSI) based on neuromuscular signals has become popular in Human-Computer Interaction. Articulatory neuromuscular activity can be captured with surface electromyography (sEMG). However, in the absence of acoustic sounds, it remains challenging to convert the articulatory neuromuscular signals of silent speakers into corresponding audio. This paper proposes a Mandarin-based SSI, a novel converter between articulatory neuromuscular from silent speakers and audio signals, by providing an auxiliary set of sEMG and audio pairs from vocal speakers. The proposed method is based on a hybrid framework combining convolutional neural networks and Transformer to utilize both global and local information. We experimentally validate the feasibility of the proposed method by obtaining an average objective character error rate (CER) of 10.69% using an automatic speech recognition (ASR) evaluation tool with four silent speakers. The results show that our Mandarin-based SSI with vocal speaker assistance facilitates the conversion from sEMG into audio in the cross-subject scenario.

US-Net: A lightweight network for simultaneous speckle suppression and texture enhancement in ultrasound images

Background:Numerous traditional filtering approaches and deep learning-based methods have been proposed to improve the quality of ultrasound (US) image data. However, their results tend to suffer from over-smoothing and loss of texture and fine details. Moreover, they perform poorly on images with different degradation levels and mainly focus on speckle reduction, even though texture and fine detail enhancement are of crucial importance in clinical diagnosis. Methods:We propose an end-to-end framework termed US-Net for simultaneous speckle suppression and texture enhancement in US images. The architecture of US-Net is inspired by U-Net, whereby a feature refinement attention block (FRAB) is introduced to enable an effective learning of multi-level and multi-contextual representative features. Specifically, FRAB aims to emphasize high-frequency image information, which helps boost the restoration and preservation of fine-grained and textural details. Furthermore, our proposed US-Net is trained essentially with real US image data, whereby real US images embedded with simulated multi-level speckle noise are used as an auxiliary training set. Results:Extensive quantitative and qualitative experiments indicate that although trained with only one US image data type, our proposed US-Net is capable of restoring images acquired from different body parts and scanning settings with different degradation levels, while exhibiting favorable performance against state-of-the-art image enhancement approaches. Furthermore, utilizing our proposed US-Net as a pre-processing stage for COVID-19 diagnosis results in a gain of 3.6% in diagnostic accuracy. Conclusions:The proposed framework can help improve the accuracy of ultrasound diagnosis.

On the origins of multiple exponence in Crow

Abstract Multiple exponence is a phenomenon in which morphemes that encode a given piece of information are realized multiple times within a single word. Unlike other members of the Siouan language family spoken across North America, Crow possesses a set of modal auxiliaries that display multiple exponence of subject person agreement. I show that one source of multiple exponence in Crow is the grammaticalization of the inflectional future from the motion verb arrive there, which brings with it its own agreement prefixes. Multiple exponence then extended to compound modals containing the inflectional future and finally to a morphologically unrelated modal – a distinct case of multiple exponence begetting additional multiple exponence. Thus, this study contributes to the broader understanding of the diachronic pathways to multiple exponence.

Uncertainty of thermal conductivity measurement at high temperatures using guarded hot plate apparatus

The guarded hot plate (GHP) is a primary method to measure the thermal conductivity (λ) of low-to-medium λ materials. Precision measurement of λ using the GHP method is challenging because the deviations from the basic assumptions in the principles of the method should be minimized, and the effect of such deviations on λ measurements must be quantified for uncertainty assessment. In the present study, the measurement uncertainty in λ of insulation materials when using a GHP apparatus was assessed at high temperatures of up to 700 °C. Both high-temperature environment and low-λ specimen make the measurement of λ difficult compared to the near-room-temperature environment or medium-λ specimen; thus, the assessed uncertainty tends to be large. A simple model that accommodates stray heat flow was described in this study and subsequently validated by measurements in various thermal configurations around the specimen. To reduce the stray heat flow, a new set of thermocouples at various parts in the GHP apparatus was specifically calibrated to accurately measure the temperature difference across the air gap before their installation, and proper auxiliary set temperatures were found in a systematic search. By employing all of these techniques, the relative expanded uncertainty (with a coverage factor of 2) in λ measurements at 200–700 °C was assessed to be 4.6%–6.8%. The present work provided a complete procedure from the measurement of λ and its uncertainty assessment for low-λ material at high temperature, where the precision measurement and uncertainty assessment of λ have been considered problematic.

A novel weighted auxiliary set matching pursuit method for glioma in Cerenkov luminescence tomography reconstruction.

Cerenkov luminescence tomography (CLT) is a promising three-dimensional imaging technology that has been actively investigated in preclinical studies. However, because of the ill-posedness in the inverse problem of CLT reconstruction, the reconstruction performance is still not satisfactory for broad biomedical applications. In this study, a novel weighted auxiliary set matching pursuit (WASMP) method was explored to enhance the accuracy of CLT reconstruction. The numerical simulations and in vivo imaging studies using tumor-bearing mice models were conducted to evaluate the performance of the WASMP method. The results of the above experiments proved that the WASMP method achieved superior reconstruction performance than other approaches in terms of positional accuracy and shape recovery. It further demonstrates that the atom selection strategy proposed in this study has a positive effect on improving the accuracy of atoms. The proposed WASMP improves the accuracy for CLT reconstruction for biomedical applications.