Sparse Sampling Research Articles

Sound field recovery based on numerical data-driven and equivalent source method

Sparse representation has been applied to realize an accurate description of the sound field from limited set of measurements. The prerequisite of this application is that sound field can be sparsely represented in a specific basis. However, most of the sparse bases are constructed by using the physical models of sound field, and are only effective for a specific category of sound sources. In this paper, the data-driven dictionary learning approach is exploited to obtain a sparse basis of sound field. Meanwhile, to reduce the difficulty and workload of the collection of data sample, the equivalent source method is utilized to collect data samples by means of simulations and thus the data samples can be generated by taking advantage of sound field properties. The performance of the sound field recovery with sparse sampling based on learned dictionary is examined and is compared with those based on other sparse bases. The result indicates that numerical data-driven model is more flexible and is not limited to a specific category of sound sources. The advantage of the proposal is presented by the results of numerical simulation and experiment.

Multikernel positional embedding convolutional neural network for photoacoustic reconstruction with sparse data.

Photoacoustic imaging (PAI) is an emerging noninvasive imaging modality that merges the high contrast of optical imaging with the high resolution of ultrasonic imaging. Low-quality photoacoustic reconstruction with sparse data due to sparse spatial sampling and limited view detection is a major obstacle to the popularization of PAI for medical applications. Deep learning has been considered as the best solution to this problem in the past decade. In this paper, we propose what we believe to be a novel architecture, named DPM-UNet, which consists of the U-Net backbone with additional position embedding block and two multi-kernel-size convolution blocks, a dilated dense block and dilated multi-kernel-size convolution block. Our method was experimentally validated with both simulated data and in vivo data, achieving a SSIM of 0.9824 and a PSNR of 33.2744dB. Furthermore, the reconstructed images of our proposed method were compared with those obtained by other advanced methods. The results have shown that our proposed DPM-UNet has a great advantage in PAI over other methods with respect to the imaging effect and memory consumption.

Prediction of Stroke Onset Time With Combined Fast High-Resolution Magnetic Resonance Spectroscopic and Quantitative T2 Mapping.

The purpose of this work is to develop a multispectral imaging approach that combines fast high-resolution 3D magnetic resonance spectroscopic imaging (MRSI) and fast quantitative T2 mapping to capture the multifactorial biochemical changes within stroke lesions and evaluate its potentials for stroke onset time prediction. Special imaging sequences combining fast trajectories and sparse sampling were used to obtain whole-brain maps of both neurometabolites (2.0 × 3.0 × 3.0 mm3) and quantitative T2 values (1.9 × 1.9 × 3.0 mm3) within a 9-minute scan. Participants with ischemic stroke at hyperacute (0-24 h, n = 23) or acute (24 h-7d, n = 33) phase were recruited in this study. Lesion N-acetylaspartate (NAA), lactate, choline, creatine, and T2 signals were compared between groups and correlated with patient symptomatic duration. Bayesian regression analyses were employed to compare the predictive models of symptomatic duration using multispectral signals. In both groups, increased T2 and lactate levels, as well as decreased NAA and choline levels were detected within the lesion (all p < 0.001). Changes in T2, NAA, choline, and creatine signals were correlated with symptomatic duration for all patients (all p < 0.005). Predictive models of stroke onset time combining signals from MRSI and T2 mapping achieved the best performance (hyperacute: R2 = 0.438; all: R2 = 0.548). The proposed multispectral imaging approach provides a combination of biomarkers that index early pathological changes after stroke in a clinical-feasible time and improves the assessment of the duration of cerebral infarction. Developing accurate and efficient neuroimaging techniques to provide sensitive biomarkers for prediction of stroke onset time is of great importance for maximizing the proportion of patients eligible for therapeutic intervention. The proposed method provides a clinically feasible tool for the assessment of symptom onset time post ischemic stroke, which will help guide time-sensitive clinical management.

Computation of a probabilistic uranium concentration map of Norway: A digital expert elicitation approach employing random forests and artificial neural networks

We compute the first probabilistic uranium concentration map of Norway. Such a map can support mineral exploration, geochemical mapping, or the assessment of the health risk to the human population. We employ multiple non-linear regression to fill the information gaps in sparse airborne and ground-borne uranium data sets. We mimic an expert elicitation by employing Random Forests and Multi-layer Perceptrons as digital agents equally qualified to find regression models. In addition to the regression, we use supervised classification to produce conservative and alarmistic classified maps outlining regions with different potential for the local occurrence of uranium concentration extremes. Embedding the introduced digital expert elicitation in a Monte Carlo approach we compute an ensemble of plausible uranium concentrations maps of Norway discretely quantifying the uncertainty resulting from the choice of the regression algorithm and the chosen parametrization of the used regression algorithms. We introduce digitated glyphs to visually integrate all computed maps and their associated uncertainties in a loss-free manner to fully communicate our probabilistic results to map perceivers. A strong correlation between mapped geology and uranium concentration is found, which could be used to optimize future sparse uranium concentration sampling to lower extrapolation components in future map updates.

Fast Brillouin optical time domain sensing based on two-dimension sparse sampling and deep learning recovery

Brillouin optical time domain sensing (BOTDS) is a technique for acquiring the Brillouin scattering spectrum (BSS) of a fiber under test (FUT) by utilizing multiple frequency scans. However, the time-consuming nature of this process limits its applicability in scenarios that require rapid measurements. In order to enhance measurement efficiency and decrease storage capacity, without compromising measurement accuracy or spatial resolution, we propose a novel fast BOTDS method based on two-dimensional (2D) sparse sampling and residual channel attention network (RCAN). The proposed method employs sparse sampling to measure the BSS signal, which is then accurately restored to the BSS with normal frequency sampling intervals by using deep learning techniques. Temperature and strain information is extracted from the restored BSS, facilitating rapid and precise measurements. The effectiveness of the proposed method has been verified through simulation and experiments. In a proof-of-concept experiment, we employed a 3000 m bend-insensitive fiber as the sensing fiber and reduced the sampling frequency by a factor of 10. Our proposed method successfully recovered the Brillouin scattering spectrum signal, resulting in a temperature measurement uncertainty of 0.25 °C. Moreover, the measurement accuracy and spatial resolution almost remained unaffected, surpassing that of interpolation-based approaches.

Flexible nonlinear inference and change-point testing of high-dimensional spectral density matrices

This paper studies a flexible approach to analyze high-dimensional nonlinear time series of unconstrained dimension based on linear statistics calculated from spectral average statistics of bilinear forms and nonlinear transformations of lag-window (i.e. band-regularized) spectral density matrix estimators. That class of statistics includes, among others, smoothed periodograms, nonlinear statistics such as coherency, long-run-variance estimators and contrast statistics related to factorial effects as special cases. Especially, we introduce the class of nonlinear spectral averages of the spectral density matrix. Having in mind big data settings, we study a sampling design which includes a sparse sampling scheme. Gaussian approximations with optimal rate are derived for nonlinear time series of growing dimension for these frequency domain statistics and the underlying lag-window (cross-) spectral estimator under non-stationarity. For change-testing (self-standardized) CUSUM statistics are examined. Further, a specific wild bootstrap procedure is proposed to estimate critical values. Simulation studies and an application to SP500 financial returns are provided in a supplement to this paper.

Tailored mechanical performance of fused filament fabricated 316L steel components through printing parameter optimization

This research focuses on Metal Material Extrusion (MMEX) using BASF Ultrafuse® Stainless Steel 316L filament, aiming to model the influence of infill parameters to optimize material usage without compromising final reliability and safety. Experimental variations in air gap, layer height, and sample orientation are analyzed for their effects on dimensional shrinkage, morphology, microstructure, and mechanical performance. The study proposes a reliable non-linear negative exponential model for data analysis, and characterization techniques such as X-ray diffraction, microscopy, and standardized mechanical testing are conducted. Results reveal a substantial influence of raster spacing, with sparse configurations demonstrating slightly reduced linear shrinkage and non-uniform pore distribution. Tensile and flexural tests highlight distinct behaviors in solid and sparse configurations, emphasizing the latter’s potential for offering advantages such as reduced weight and cost savings, and setting sparse samples as a viable alternative for designs with extensive bed-parallel surfaces. The findings contribute to the development of novel design-for-manufacturing strategies for functional, structural elements using MMEX and offer insights for optimizing the fabrication process by understanding the role of raster deposition.

Visualization and quantitative evaluation of delamination defects in GFRPs via sparse millimeter-wave imaging and image processing

Multilayer glass fiber reinforced polymer (GFRP) has broad applications, playing a pivotal role in various fields including aerospace and wind energy production. These structures may develop internal delamination defects during their manufacture or service, which can significantly impair the integrity and safety of the equipment. Thus, the inspection of such delamination defects during the service life of the equipment is crucial for assessing its operational performance and lifespan. Millimeter-wave (MMW) nondestructive testing (NDT) presents a promising method for detecting these defects in GFRPs, due to its high resolution and powerful penetration characteristics. This paper proposes a visual NDT and quantitative characterization method based on sparse imaging and image processing. This method encompasses the following three key points. (i) To facilitate rapid detection and visualization in monostatic mode, we proposed a detection approach utilizing sparse scanning and imaging. Initially, we developed a sparse sampling control matrix grounded to guide the probe for spatial automatic sparse sampling, thereby enhancing data acquisition efficiency and reducing data volume. Subsequently, we introduced a rapid sparse imaging method, integrating wave spectrum reconstruction and compressed sensing (CS). Beginning from the original data domain, we designed a fast imaging operator employing forward and inverse transformations of spectrum reconstruction imaging. (ii) In order to further refine the quality of the detection images, we proposed an adaptive singular value decomposition (SVD) approach to suppress the coupling signal and enhance the specimen signal. By adaptively adjusting the singular values, we generated a signal with the optimal target-to-background ratio (TBR). (iii) We ascertained the position and area parameters of the delamination defects using the proposed defect quantification method, combining image segmentation and image statistics. Experimental validation affirmed the effectiveness of the proposed method, revealing the following results: 1) The proposed method, despite utilizing a small amount of observational data, generates better visual images while maintaining efficient detection; 2) It can effectively identify all 0.5 mm-thick internal delamination defects in GFRPs, simultaneously retaining high quantitative characterization precision; 3) The positional quantification error remains less than 0.5 mm, and the relative area quantification error is constrained within 8%.

Parametric Reduced Order Model of a Gas Bearings Supported Rotor

Abstract Gas bearings use pressurized gas as a lubricant to support and guide rotating machinery. These bearings have a number of advantages over traditional lubricated bearings, including higher efficiency in a variety of applications and reduced maintenance requirements. However, they are more complex to operate and exhibit nonlinear behaviors. This paper presents a parametric hyper reduced order model (h-ROM) of a gas bearings supported rotor enabling to speed up the computations up to a factor 100 while preserving satisfactory accuracy. A Galerkin projection setting is employed to reduce the dimension of the governing equations and the nonlinear terms are efficiently tackled with a sparse sampling technique. The performances of the h-ROM are compared to a high fidelity model both in terms of accuracy and computation time, demonstrating the potential for future anomaly detection applications.

Development and internal validation of a clinical prediction model for individualized dosing of BAY 81-8973, A full-length recombinant factor VIII, in pediatric patients with haemophilia A

BackgroundAs the most commonly used coagulation factor VIII (FVIII) concentrate in China, the individualized dosing prediction model of Kovaltry (BAY81–8973) is not fully investigated in pediatric patients. The prophylaxis tailored by population pharmacokinetic (PopPK) model can optimize dosing regimens. ObjectivesThis study aimed to develop PopPK models of BAY 81–8973 in pediatric patients, identify quantitative relationships of blood type (as a substitution for von Willebrand factor) on FVIII clearance and provide model-informed precision dosing (MIPD) procedures. MethodsPediatric patients with severe hemophilia A were enrolled and PK tests were conducted. The blood samples were collected at six time point. One-stage-based activated partial thromboplastin time was used for FVIII activity. Basic demographics and key covariates (blood type and von Willebrand factor antigen) were collected. A nonlinear mixed-effect modeling (NONMEM) approach was employed to establish PopPK model. Simulations were performed to evaluate current dosing regimens and present MIPD strategies. ResultsA total of 30 pediatric patients were included in the analysis. In the final model, Fat-free mass calculated from weight, age, and height was included as a size descriptor which affect FVIII apparent volume of distribution and clearance. Both von Willebrand factor antigen (VWF:Ag) and blood type accounted for the interindividual variability of FVIII clearance, but only one can retain in the final model. Therefore, two PopPK models based on VWF:Ag or blood type were developed. When the VWF:Ag value is doubled, the FVIII clearance is reduced by 35 %. Compared with blood type non-O, the clearance in pediatric patients with blood type O increased by 25.9 %. Weight-based dosing without regard to age and blood type resulted in large differences in FVIII trough activity. Patient demographics, dosing information, sparse blood samples and PopPK model together with Bayesian estimate constituted the MIPD workflow. Using it, the individual parameters and optimized dosing regimen could be achieved. ConclusionThis is the first predictive model designed to predict individualized dosing of BAY 81–8973 in pediatric patients with Hemophilia A. These results are useful in the PK-guided prophylaxis among pediatric patients and hold great potential to improve their long-term clinical outcomes.

Robustness of baryon acoustic oscillations measurements with photometric redshift uncertainties

ABSTRACT We investigate the robustness of baryon acoustic oscillations (BAO) measurements with a photometric galaxy sample using mock galaxy catalogues with various sizes of photometric redshift (photo-z) uncertainties. We first conduct the robustness of BAO measurements, assuming we have a perfect knowledge of photo-z uncertainties. We find that the BAO shift parameter α can be constrained in an unbiased manner even for 3 per cent photometric redshift uncertainties up to z ∼ 1. For instance, α = 1.006 ± 0.078 with 95 per cent confidence level is obtained from 3 per cent photo-z uncertainty data at z = 1.03 using the sample of M* ≥ 1010.25 M⊙ h−2. We also find that a sparse galaxy sample, e.g. &amp;lt;2 × 10−4 [h Mpc−1]3, causes additional noise in the covariance matrix calculation and can bias the constraint on α. Following this, we look into the scenario where incorrect photometric redshift uncertainties are assumed in the fitting model. We find that underestimating the photo-z uncertainty leads to a degradation in the constraining power on α. However, the constrained value of α is not biased. We also quantify the constraining power on Ωm0 assuming the Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST)-like covariance and find that the 95 per cent confidence level is σ(Ωm0) ∼ 0.03–0.05 corresponding to the photo-z uncertainties of 1–3 per cent, respectively. Finally, we examine whether the skewness in the photometric redshift can bias the constraint on α and confirm that the constraint on α is unbiased, even assuming a Gaussian photo-z uncertainty in our model.

A fusion method for capacity estimation of fast-charging lithium-ion battery under varied degradation mode

Recently, lithium-ion batteries with fast-charging capability have been gradually equipped in electric products. Capacity estimation specially developed for fast-charging batteries is still an open question, and most of the existing works are proposed for batteries charged under the 1C rate. However, incorrect fast-charging strategies may result in Li plating leading to battery capacity aging rapidly and degradation mode variation. Moreover, unprecise sampling by BMS increases the difficulty of accurately estimating battery capacity in such scenarios. Thus, a fusion prognostic method based on ensemble learning is proposed for the above issues in fast-charging batteries. Firstly, measurement-based health features are extracted from the portion charging phase. Then, the accuracy of the validation dataset-based strategy is proposed to achieve time-varying weight allocation. Finally, a fusion prognostic model is constructed based on the above weight allocation strategy and health features. The effectiveness of the proposed fusion method is verified by a self-design fast-charging battery dataset that maintains stable and reliable performance under sparse sampling and degradation mode variation conditions. In addition, the robustness and adaptability are validated by the comparison experiments with traditional ensemble learning-based methods, in which accuracy is improved by 27.5% minimally.

MJAR: A novel joint generalization-based diagnosis method for industrial robots with compound faults

Compound faults inevitably occur in multi-joint industrial robots resulting in excessive vibration. Intelligent diagnosis for the occurrence and position of fault joints can efficiently reduce the maintenance cost. Compared with laboratory disassembly parts, joint diagnosis of in-situ industrial robots is more challenging due to coupling of multiple subsystems and the industrial noise interference. This paper proposes a novel joint generalization-based fault diagnosis method for industrial robots via the multi-joint attention residual network (MJAR) model. Each joint signals are independently input to parallel residual convolution with unidirectional matrix kernel (ResCUM) in multi-joint decoupling attention (MJDA) module, which is exclusively designed combining residual and attention mechanism to provide decoupling capability of compound faults. And the multi-spatial reconstruction (MSR) module based on sparse sampling is design to provide multi-scale feature extraction adapted to real industrial signals. MJAR can diagnose unseen compound fault combinations without using any transfer strategy as the robot joint generalization-based fault diagnosis (RJGFD) framework. Integrated experiment results show that the MJAR is superior to some advanced congener DL methods.

Weighted Dynamic Time Warping for Time Series

Recurrence network is a typical time series analysis method. However, irregular sampling may overshadow the dynamic features characterized by traditional recurrence network method, which makes the method ineffective. This paper introduces dynamic time warping method to determine the distance between time series segments. The method can match the features of time series segments and reduce the distortion of distances caused by irregular sampling. For points in sparse sampling area, many characteristics may not be collected and thus the matching results are unreliable. To overcome this problem, we further propose weighted dynamic time warping method which assigns small weights to matching points with sparse local sampling so as to reduce their influence on distance. A recurrence network constructed by weighted dynamic time warping method can effectively capture underlying nonlinear and nonstationary system dynamics from irregular sampling time series. The effectiveness of the proposed method is illustrated by two case studies: the discrimination of different dynamic behaviors and detection of system parameter perturbation from time series of Rössler system.

Large Interferometer For Exoplanets (LIFE)

Context. A mid-infrared nulling-space interferometer offers a promising way to characterize thermal light from habitable planet candidates around Sun-like stars. However, one of the main challenges inherent in achieving this ambitious goal is the high-precision stability of the optical path difference and amplitude over a few days for planet detections and all the way up to a few weeks for in-depth characterization. This is related to mission parameters such as aperture size, number of apertures, and total instrument throughput. Aims. Here, we propose a new method called phase-space synthesis decomposition (PSSD) to shorten the stability requirement to a scale of minutes, significantly relaxing the technological challenges of the mission. Methods. By focusing on the consideration of what exactly modulates the planetary signal in the presence of the stellar leak and systematic error, PSSD prioritizes the modulation of the signals along the wavelength domain rather than baseline rotation. Modulation along the wavelength domain allows us to extract source positions in parallel to the baseline vector for each exposure. The sum of the one-dimensional data is converted into two-dimensional information. Based on the reconstructed image, we constructed a continuous equation and extract the spectra through the singular value decomposition, while efficiently separating them from a long-term systematic stellar leak. Results. We performed numerical simulations to investigate the feasibility of PSSD for the Large Interferometer For Exoplanets (LIFE) mission concept. We confirm that multiple terrestrial planets in the habitable zone around a Sun-like star at 10 pc can be detected and characterized despite high levels and long durations of systematic noise. We also find that PSSD is more robust against a sparse sampling of the array rotation compared to purely rotation-based signal extraction. Using PSSD as signal extraction method significantly relaxes the technical requirements on the signal stability and further increases the feasibility of the LIFE mission.

Joint DOA and Frequency Estimation With Spatial and Temporal Sparse Sampling Based on 2-D Covariance Matrix Expansion

Joint DOA and Frequency Estimation With Spatial and Temporal Sparse Sampling Based on 2-D Covariance Matrix Expansion

Improving Dose Volume Histogram (DVH) Based Analysis of Clinical Outcomes Using Modern Statistical Techniques: A Systematic Answer to Multiple Comparisons Concerns

DVH constraints are essential in the clinical practice of radiation therapy. Historically, DVH constraints were found through sparse sampling of all possible DVH indices to find one that appeared to be most predictive for clinical toxicity. This approach can lead to inconsistent results among studies and to multiple comparison concerns. We aim to solve both problems by examining a full array of DVH indices using statistical methods that account for strong correlations among DVH indices and incorporate radiobiological knowledge constraints. We extracted a dense array of V%_D indices from a treatment planning system using ESAPI interface, with V%_D corresponding to the volume fraction irradiated to dose D, or higher. We used Fused Lasso as the base model to compensate for correlations among DVH indices because it applies a penalty on the difference between DVH variables with adjacent dose. The base model was augmented with additional constraints based on radiobiological considerations: the positivity constraint (beta_i > 0) which assumes that any tissue irradiation cannot reduce the risk of toxicity, and monotonicity constraint (beta_i+1 > = beta_i) which assumes that higher dose to a fixed volume fraction cannot be associated with a lower risk of toxicity. We called the hybrid model KC-Lasso (Knowledge Constrained Lasso) and applied it to two clinical examples: grade 2 acute rectal toxicity in conventionally fractionated RT for 79 prostate cancer patients (77.4 Gy + MR based boost to 81-83 Gy) and cardiac toxicity in conventionally fractionated RT for 119 locally advanced Non-small Cell Lung Cancer (NSCLC) patients (Median prescribed dose 62 Gy). We further examined alternative data driven models to determine the importance of knowledge constraints. KC-Lasso detected two distinct dose thresholds for grade 2 rectal toxicity, at 35 Gy and 78 Gy. A threshold of 51 Gy was detected for reduced overall survival due to cardiac irradiation in NSCLC patients. An examination of KC-Lasso models at varying step size suggested that a single mid-range index can be used as a treatment planning constraint while full model can be used for confirmatory, final plan evaluation. Alternative models which lack knowledge constraints show patterns of negative and isolated coefficients which are difficult to interpret and are not likely to be generalizable. A more systematic approach to the analysis of correlations between DVH constraints and clinical toxicity can lead to greater consistency of results among different studies, better understanding of true dose thresholds and results which are more generalizable.

Real Time Volumetric MRI for MR-Guided 3D Motion Tracking via Sparse Prior-Augmented Neural Representation Learning

To reconstruct volumetric MRI from orthogonal cine acquisition aided by sparse priors of 2 static 3D MRI through implicit neural representation (NeRP) learning, with the goal of eliminating large-scale training datasets for data-driven sparse MRI reconstruction and supporting clinical workflow of real time 3D motion tracking during MR-guided radiotherapy. A multi-layer perceptron network was trained to learn the NeRP of a patient-specific MRI dataset, where the network takes 4D data coordinates of voxel locations and motion states as inputs and outputs corresponding voxel intensities. By first learning the NeRP of 2 static 3D MRI with different breathing motion states, prior knowledge of patient breathing motion was embedded into network weights through optimization. The prior knowledge was then augmented from 2 to 31 motion states by querying the optimized network at interpolated/extrapolated motion state coordinates. Starting from the prior-augmented network as an initialization point, the network was further trained using sparse samples of 2 orthogonal cine slices. The final volumetric reconstruction was obtained by querying the trained network at desired 3D spatial locations. We evaluated the proposed method using 5-minute volumetric MRI time series with 340 ms temporal resolution collected from 7 liver carcinoma patients. The time series was acquired using golden-angle radial MRI sequence and reconstructed through retrospective sorting. Two MRI with inhale and exhale states respectively were selected from the first 30 sec of the time series for prior embedding and augmentation. The remaining 4.5-min time series was used for volumetric reconstruction evaluation, where we retrospectively subsampled each MRI to 2 orthogonal slices and compared network-reconstructed images to ground truth images in terms of image quality and the capability of supporting 3D target motion tracking. Across the 7 patients evaluated, the peak signal to noise ratio between model reconstruction and ground truth was 54.66 ± 6.16 dB and the structural similarity index measure was 0.99 ± 0.01. Gross tumor volume (GTV) contours estimated by deforming a reference state MRI to model-reconstructed and ground truth MRI showed good consistency. The 95-percentile Hausdorff distance between GTV contours was 1.89 ± 1.13 mm, which is less than the voxel dimension. The mean GTV centroid position difference between ground truth and model estimation was less than 1 mm in all 3 orthogonal directions. Volumetric MRI from orthogonal cine acquisition with sparse priors is feasible by modeling prior knowledge through implicit neural representation learning. The model-reconstructed images showed sufficient accuracy in supporting 3D motion tracking of abdominal targets. By eliminating the need for large scale training datasets, the method promises to enable clinical implementation of 3D motion tracking for precision radiation therapy.

MMTSA

Multimodal sensors provide complementary information to develop accurate machine-learning methods for human activity recognition (HAR), but introduce significantly higher computational load, which reduces efficiency. This paper proposes an efficient multimodal neural architecture for HAR using an RGB camera and inertial measurement units (IMUs) called Multimodal Temporal Segment Attention Network (MMTSA). MMTSA first transforms IMU sensor data into a temporal and structure-preserving gray-scale image using the Gramian Angular Field (GAF), representing the inherent properties of human activities. MMTSA then applies a multimodal sparse sampling method to reduce data redundancy. Lastly, MMTSA adopts an inter-segment attention module for efficient multimodal fusion. Using three well-established public datasets, we evaluated MMTSA's effectiveness and efficiency in HAR. Results show that our method achieves superior performance improvements (11.13% of cross-subject F1-score on the MMAct dataset) than the previous state-of-the-art (SOTA) methods. The ablation study and analysis suggest that MMTSA's effectiveness in fusing multimodal data for accurate HAR. The efficiency evaluation on an edge device showed that MMTSA achieved significantly better accuracy, lower computational load, and lower inference latency than SOTA methods.

Reciprocating compressor health monitoring based on BSInformer with deep convolutional AutoEncoder

A composite model combining deep convolutional autoencoder and balanced sparse sampling Informer (MCA-BSInformer) is proposed to solve the problem that reciprocating compressors are difficult to achieve long-term accurate health monitoring. By introducing the Maximum Mean Discrepancy (MMD) difference term into the deep convolutional autoencoder, the distribution of noise data in the multi-source signal of the compressor become more similar to the distribution of normal training data. Thus, effective low-dimensional features containing spatial information in the data can be obtained to reduce the impact of noise data. In addition, a balanced sparse sampling algorithm is integrated into the Informer model to achieve adaptive sparse sampling. The improved Informer model effectively solves the problem that the model reduces the prediction performance while pursuing efficiency. The experimental results demonstrate that the prediction accuracy of the composite model for the health status of reciprocating compressors is higher than that of other mainstream models.