Frequency Domain Research Articles

Semi-supervised segmentation of abdominal organs and liver tumor: uncertainty rectified curriculum labeling meets X-fuse

Precise liver tumors and associated organ segmentation hold immense value for surgical and radiological intervention, enabling anatomical localization for pre-operative planning and intra-operative guidance. Modern deep learning models for medical image segmentation have evolved from convolution neural networks to transformer architectures, significantly boosting global context understanding. However, accurate delineation especially of hepatic lesions remains an enduring challenge due to models’ predominant focus solely on spatial feature extraction failing to adequately characterize complex medical anatomies. Moreover, the relative paucity of expertly annotated medical imaging data restricts model exposure to diverse pathological presentations. In this paper, we present a three-phrased cascaded segmentation framework featuring an X-Fuse model that synergistically integrates spatial and frequency domain’s complementary information in dual encoders to enrich latent feature representation. To enhance model generalizability, building upon X-Fuse topology and taking advantage of additional unlabeled pathological data, our proposed integration of curriculum pseudo-labeling with Jensen–Shannon variance-based uncertainty rectification promotes optimized pseudo supervision in the context of semi-supervised learning. We further introduce a tumor-focus augmentation technique including training-free copy-paste and knowledge-based synthesis that show efficacy in simplicity, contributing to the substantial elevation of model adaptability on diverse lesional morphologies. Extensive experiments and modular evaluations on a holdout test set demonstrate that our methods significantly outperform existing state-of-the-art segmentation models in both supervised and semi-supervised settings, as measured by the Dice similarity coefficient, achieving superior delineation of bones (95.42%), liver (96.26%), and liver tumors (89.53%) with 16.41% increase comparing to V-Net on supervised-only and augmented-absent scenario. Our method marks a significant step toward the realization of more reliable and robust AI-assisted diagnostic tools for liver tumor intervention. We have made the codes publicly available [https://github.com/lyupengju/X-Fuse].

Forecasting Inflation with the New Keynesian Phillips Curve: Frequencies Matter*

AbstractWe forecast US inflation with a new Keynesian Phillips curve (NKPC) in the frequency domain. Our method consists of decomposing the time series of inflation and its NKPC predictors into several frequency bands, forecasting separately each frequency component of inflation, and then summing up those forecasts to obtain the forecast for aggregate inflation. We find that (i) accurately forecasting the low frequency of inflation is, on average, crucial to successfully forecast inflation; (ii) our NKPC low‐frequency forecast model consistently and significantly outperforms the time‐series NKPC and standard benchmark models; (iii) the low frequencies of inflation expectations and unemployment are the key predictors; and (iv) optimally switching on / off the forecasts of each frequency components of inflation at each period allows to outstandingly track inflation and show that all frequencies of inflation matter.

Modelling Li-V2O5 Batteries Using Galvanostatic Intermittent Titration Technique and Electrochemical Impedance Spectroscopy: Towards Final Applications

Given the relevance of lithium and post-lithium batteries as electrochemical energy storage systems, the peculiar crystalline structure of V2O5 and its doping capacity play key roles in lithium-ion battery technology. To integrate them in high-efficiency modules, systematic methodologies are required to estimate the state of charge in a reliable way and predict the Li-V2O5 battery’s performance according to their electrochemical phenomena, including two plateaus in the galvanostatic cycling curves and the dynamic behavior governed by the energy demand. Most state of charge estimation and battery modeling procedures are focused on conventional Li-batteries that show a unique plateau. In this work, we propose a systematic methodology based on the galvanostatic intermittent titration technique and electrochemical impedance spectroscopy to study battery performance in the time and frequency domains, respectively. The proposed methodology, with a time–frequency correlation, promotes a deeper understanding of the electrochemical phenomena and general behavior of Li-V2O5 batteries, allowing for its subsequent extrapolation to more complex and higher-capacity lithium and post-lithium batteries used in high-power applications with a minimum error.

Life prediction method of rolling bearing based on CNN-LSTM-AM

Bearing is the key component to determine the health of machinery, and it is of great significance to monitor its working status in real time and predict its remaining useful life. In recent years, the RUL prediction method based on deep learning has been widely used and achieved good prediction results. Here, a bearing life prediction method based on convolution neural network (CNN), long short term memory (LSTM) and attention mechanism (AM) is proposed. First of all, the time domain and frequency domain features of the original vibration signals of rolling bearings are extracted, and the extracted feature set is normalized as the input of CNN. The main function of CNN is to extract spatial features and reduce the dimension of the data. Then, using LSTM to extract the information that may be ignored by CNN, the feature information extracted by CNN-LSTM is input to the attention mechanism for weighting, and the key information is screened. And then more accurately represent the degradation characteristics of the equipment, and finally get the bearing remaining life. The performance of the model is verified by two sets of public data sets, and the experimental results show that it is compared with the CNN-LSTM method. The root mean square error (RMSE) index based on CNN-LSTM-AM method is reduced by 14.6 % and 13.8 % respectively, and the score index is increased by 2.0 % and 2.4 % respectively. The results show that the proposed method has higher accuracy in bearing RUL prediction.

Distributed temperature sensing on chip by optical frequency domain reflectometer

Distributed temperature sensing on chip by optical frequency domain reflectometer

Fast and accurate methodologies for the eddy current losses analysis in axially laminated rotor of synchronous reluctance machines under different loads

AbstractThe authors present two fast and accurate methodologies for the computation of eddy current losses in the axially laminated rotor of a synchronous reluctance machine. The methodologies are based on different combinations of the finite element method in time and frequency domains with 2D and 3D formulations. First, a comparative study of the 2D and 3D formulations for loss calculation is presented, considering various load angles of the machine to illustrate the problem of eddy‐current losses in this type of machines and its dependence on the load angle. The influence of the iron saturation on the loss calculation is also evaluated in these computations. A novel correction factor based on the computations at two load angles is proposed to convert the losses computed from a 2D model to match those computed from a 3D model. For the sake of generality, investigations are also conducted for various thicknesses of the lamination layers and different machine lengths, and an analytical method to describe the dependency of eddy‐current losses on the load angle of the machine is introduced. Moreover, a simplified method is proposed for modelling eddy currents in the frequency domain and calculating losses in an axially laminated structure based solely on the results of a magnetostatic solution. The results obtained by the simplified model demonstrate excellent agreement with the full 3D magneto‐dynamic simulation. Overall, the findings contribute to understanding and accurately characterising the eddy current losses in axially laminated rotors, offering potential insights for designing and optimising axially laminated synchronous reluctance electric machines.

Advancing Hyperspectral and Multispectral Image Fusion: An Information-Aware Transformer-Based Unfolding Network.

In hyperspectral image (HSI) processing, the fusion of the high-resolution multispectral image (HR-MSI) and the low-resolution HSI (LR-HSI) on the same scene, known as MSI-HSI fusion, is a crucial step in obtaining the desired high-resolution HSI (HR-HSI). With the powerful representation ability, convolutional neural network (CNN)-based deep unfolding methods have demonstrated promising performances. However, limited receptive fields of CNN often lead to inaccurate long-range spatial features, and inherent input and output images for each stage in unfolding networks restrict the feature transmission, thus limiting the overall performance. To this end, we propose a novel and efficient information-aware transformer-based unfolding network (ITU-Net) to model the long-range dependencies and transfer more information across the stages. Specifically, we employ a customized transformer block to learn representations from both the spatial and frequency domains as well as avoid the quadratic complexity with respect to the input length. For spatial feature extractions, we develop an information transfer guided linearized attention (ITLA), which transmits high-throughput information between adjacent stages and extracts contextual features along the spatial dimension in linear complexity. Moreover, we introduce frequency domain learning in the feedforward network (FFN) to capture token variations of the image and narrow the frequency gap. Via integrating our proposed transformer blocks with the unfolding framework, our ITU-Net achieves state-of-the-art (SOTA) performance on both synthetic and real hyperspectral datasets.

Hot rolled steel surface defect detection and classification using an automatic ensemble approach

This study introduces an ensemble-based Deep Neural Network (DNN) model for detecting defects on steel surfaces. The method suggested in this study classifies steel surface conditions into six possible fault categories, namely, crazing, inclusion, rolled in, pitted surface, scratches, and patches. The images undergo preprocessing and extraction of features in spatial and frequency domains using image segmentation techniques such as grey level difference method (GLDM), fast Fourier Transform (FFT), grey level co-occurrence matrix (GLCM), texture analysis and discrete wavelet transform (DWT). The ensembling of image features into a fused feature pool is carried out after the preprocessing of input images that are provided as input to a light-weight neural network model for training and testing. The performance of the model is comprehensively evaluated via an ablation study both before and after ensembling. In addition, the model capability is effectively analyzed using receiver operating characteristics (ROC) curve, confusion matrix from which classification accuracy of the model could be obtained and other parameters including precision and f1-score. It was observed that the proposed deep learning network presents phenomenally high accuracy of 99.72% for detection and classification of steel surface faults. This result was found to be superior when compared with the performance of the same neural network over each feature type individually. This study also compares the classification results of the model built based on the ensembled feature set with the results of various other classification approaches available in literature. The ensemble-based model could potentially be integrated into existing inspection systems for real-time, efficient and robust condition monitoring of steel surfaces.

WBUN: an interpretable convolutional neural network with wavelet basis unit embedded for fault diagnosis

Convolutional Neural Network (CNN) is extensively applied in mechanical system fault diagnosis. However, the absence of transparent decision mechanisms in CNNs hinders credibility. To address these challenges, this paper proposes an interpretable wavelet basis unit convolutional network (WBUN). This network incorporates meticulously designed wavelet basis unit (WBU) functions into convolutional layer, creating the interpretable wavelet basis unit convolutional (WBUConv) layer. Convolutional kernels with clear physical significance enable the WBUConv layer to extract fault-related features in both time and frequency domains, enhancing diagnostic performance, and interpreting the CNN’s attention frequency along with the convolutional kernel’s training outcomes. In this paper, three WBU functions are designed to construct the corresponding WBUNs, and their effectiveness and interpretability are verified through three sets of mechanical fault diagnosis experiments. Meanwhile, experimental results demonstrate the WBUConv layer’s remarkable advantages in noise robustness, convergence speed, and strong generalization ability.

Interferometric-scale full-field vibration measurement by a combination of digital image correlation and laser vibrometer

Digital Image Correlation (DIC) is a crucial noncontact full-field optical measurement method used in various fields. However, in practical applications, DIC is affected by systematic and random noises, leading to experimental resolutions lower than theoretical ones. In this study, we proposed a laser Doppler vibrometer guided DIC to perform vibration measurements. 3D-DIC obtains a sequence of out of displacement field initially. A three-dimensional frequency domain collaborative filtering (3D-FDCF) method that utilizes Laser Doppler vibrometer (LDV) single-point data to assist in processing of the displacement field sequence pixel-wise is adopted. The 3D-FDCF method combines lowpass filtering in the spatial frequency domain with LDV-guided bandpass filtering in the temporal frequency domain. The effectiveness of the 3D-FDCF method is demonstrated through a comparison among DIC data, the filtered DIC data, and continuously scanning LDV data. The experiment results demonstrate the 3D-FDCF method's capability in measuring vibration amplitudes of several hundred nanometers with the size of a test sample of about 100 mm × 100 mm, supporting the statement of interferometric scale full-field vibration measurement by DIC with the guidance of the LDV data.

Frequency domain stability and relaxed convergence conditions for filtered error adaptive feedforward

Frequency domain stability and relaxed convergence conditions for filtered error adaptive feedforward

ANALYSIS OF THE BIFURCATING DUCT OF AN INLET PARTICLE SEPARATOR IN TRANSONIC FLOW CONDITIONS

Abstract To increase the reliability of turboprop and turboshaft engines in extreme operating conditions, filtering protections such as Inlet Particle Separators (IPS) can be installed at the intake. The flow inside an IPS is highly 3D and unsteady, with fluctuations especially pronounced when transonic conditions are reached. In this paper, we aim at providing the transonic analysis of an industrial IPS designed for aerodynamic lab testing. As a first step, we define the threshold beyond which sonic conditions are reached in the boundary cross sections of the IPS by means of a 1D semi-empirical model. Second, we select a critical configuration, and we perform CFD simulations to analyze the locations and evolution of the transonic flow inside the IPS. Different models are presented, i.e. steady and unsteady Reynolds-Averaged Navier-Stokes, Detached Eddy Simulation and Large Eddy Simulation, characterized by three levels of resolution. The results show the evolution of some transonic shocks: the steady-state model is only providing information on averaged quantities, while the time-resolved simulations offer a more precise overview in the time domain. The analysis in the frequency domain reveals the frequencies of the transonic fluctuations. Finally, we discuss three alternative designs to effectively improve the operating range and mitigate the risks related to the transonic instabilities, comparing the differences in separation efficiency with respect to the baseline case. The results prove that the optimal design choice is given by the trade-off between operating range, pressure losses and separation efficiency.

Time distribution of stimulated transition probabilities

In recent years, an increasing number of experimental results have been published for a wide variety of systems in which high-precision, ultra-fast, and ultra-short laser pulses techniques have been used to study time-resolved dynamical processes. In the absence of a theoretical prediction of the time distribution of the transition probabilities, the lifetimes τ are generally extracted by fitting the decay time with an exponential, bi-exponential or three-exponential function. In fitting the data, the short-time behavior is generally neglected. The purpose of this study is to show that an explicit formula for the time distribution of the transition probability can be determined rigorously using time-dependent perturbation theory. A formula that perfectly fits an important subset of the experimental results from t = 0 to t=∞ . We show that by following a route different from the usual procedure for deriving transition amplitudes and Einstein coefficients in time-dependent perturbation theory, one ends up with a time-dependent factor, that is, a temporal distribution function of the transition probabilities, and for the Einstein coefficients. The time distribution reported here looks in the time domain similar to the Planck distribution in the frequency domain. In fact, the behavior of the time distribution, with a maximum at 2τln⁡2 , and with τ playing the role of temperature, resembles the behavior of the Planck frequency distribution. We present several examples to demonstrate that the theoretical formula allows for easier fitting of the experimental results reported in the literature. We also show that the time distribution explains the difference in resonance heights between the experimental and theoretical blue laser spectra.

Eduction of coherent structures from schlieren images of twin jets using SPOD informed with momentum potential theory in the spectral domain

Abstract This work presents a methodology to extract coherent structures from high-speed schlieren images of turbulent twin jets which are more physically interpretable than those obtained with currently existing techniques. Recently, Prasad and Gaitonde (J Fluid Mech 940:1–11, 2022) introduced an approach which employs the momentum potential theory of Doak (J Sound Vib 131(1):67–90, 1989) to compute potential (acoustic and thermal) energy fluctuations from the schlieren images by solving a Poisson equation, and combines it with spectral proper orthogonal decomposition (SPOD) to educe coherent structures from the momentum potential field instead of the original schlieren field. While the latter field is dominated by a broad range of vortical fluctuations in the turbulent mixing region of unheated high-speed jets, the momentum potential field is governed by fluctuations which are intimately related to acoustic emission, and its spatial structure in the frequency domain is very organized. The proposed methodology in this paper improves the technique of Prasad and Gaitonde (J Fluid Mech 940:1–11, 2022) in three new ways. First, the solution of the Poisson equation is carried out in the frequency-wavenumber domain instead of the time-space domain, which simplifies and integrates the solution of the Poisson equation within the SPOD framework based on momentum potential fluctuations. Second, the issue of solving the Poisson equation on a finite domain with ad hoc boundary conditions is explicitly addressed, identifying and removing those unphysical harmonic components introduced in the solution process. Third, the solution of the SPOD problem in terms of momentum potential fluctuations is used to reconstruct schlieren SPOD fields associated with each mode, allowing the visualization of the obtained coherent structures also in terms of the density gradient. The method is applied here to schlieren images of a twin-jet configuration with a small jet separation at two supersonic operation conditions: a perfectly-expanded and an overexpanded one. The SPOD modes based on momentum potential fluctuations retain the wavepacket structure including the direct Mach-wave radiation, together with upstream- and downstream-traveling acoustic waves, similar to SPOD modes based on the schlieren images. However, for the same dataset, they result in a lower-rank decomposition than schlieren-based SPOD and provide an effective separation of twin-jet fluctuations into independent toroidal and flapping oscillations that are recovered as different SPOD modes. These coherent structures are more consistent with twin-jet wavepacket models available in the literature than those originally obtained with direct schlieren-based SPOD, facilitating their interpretation and comparison against theoretical analyses. Graphical abstract

Analytically separating the source of the Teukolsky equation

Recent gravitational wave detections from black hole mergers have underscored the critical role black hole perturbation theory and the Teukolsky equation play in understanding the behavior of black holes. The separable nature of the Teukolsky equation has long been leveraged to study the vacuum linear Teukolsky equation; however, as theory and measurements advance, solving the sourced Teukolsky equation is becoming a frontier of research. In particular, second-order calculations, such as in quasinormal mode and self-force problems, have extended sources. This paper presents a novel method for analytically separating the Teukolsky equation’s source, aimed to improve efficiency. Separating the source is a nontrivial problem due to the angular and radial mixing of generic quantities in Kerr spacetime. We provide a proof-of-concept demonstration of our method and show that it is accurate, separating the Teukolsky source produced by the stress-energy tensor of an ideal gas cloud surrounding a Kerr black hole. The detailed application of our method is provided in an accompanying notebook. Our approach opens up a new avenue for accurate black hole perturbation theory calculations with sources in both the time and frequency domain. Published by the American Physical Society 2024

A Comprehensive Overview of Network Slicing for Improving the Energy Efficiency of Fifth-Generation Networks.

The introduction of fifth-generation (5G) mobile networks leads to an increase in energy consumption and higher operational costs for mobile network operators (MNOs). Consequently, the optimization of 5G networks' energy efficiency is crucial, both in terms of reducing MNO costs and in terms of the negative environmental impact. However, many aspects of the 5G mobile network technology itself have been standardized, including the 5G network slicing concept. This enables the creation of multiple independent logical 5G networks within the same physical infrastructure. Since the only necessary resources in 5G networks need to be used for the realization of a specific 5G network slice, the question of whether the implementation of 5G network slicing can contribute to the improvement of 5G and future sixth-generation networks' energy efficiency arises. To tackle this question, this review paper analyzes 5G network slicing and the energy demand of different network slicing use cases and mobile virtual network operator realizations based on network slicing. The paper also overviews standardized key performance indicators for the assessment of 5G network slices' energy efficiency and discusses energy efficiency in 5G network slicing lifecycle management. In particular, to show how efficient network slicing can optimize the energy consumption of 5G networks, versatile 5G network slicing use case scenarios, approaches, and resource allocation concepts in the space, time, and frequency domains have been discussed, including artificial intelligence-based implementations of network slicing. The results of the comprehensive discussion indicate that the different implementations and approaches to network slicing pave the way for possible further reductions in 5G MNO energy costs and carbon dioxide emissions in the future.

Integrating gated recurrent unit in graph neural network to improve infectious disease prediction: an attempt

ObjectiveThis study focuses on enhancing the precision of epidemic time series data prediction by integrating Gated Recurrent Unit (GRU) into a Graph Neural Network (GNN), forming the GRGNN. The accuracy of the GNN (Graph Neural Network) network with introduced GRU (Gated Recurrent Units) is validated by comparing it with seven commonly used prediction methods.MethodThe GRGNN methodology involves multivariate time series prediction using a GNN (Graph Neural Network) network improved by the integration of GRU (Gated Recurrent Units). Additionally, Graphical Fourier Transform (GFT) and Discrete Fourier Transform (DFT) are introduced. GFT captures inter-sequence correlations in the spectral domain, while DFT transforms data from the time domain to the frequency domain, revealing temporal node correlations. Following GFT and DFT, outbreak data are predicted through one-dimensional convolution and gated linear regression in the frequency domain, graph convolution in the spectral domain, and GRU (Gated Recurrent Units) in the time domain. The inverse transformation of GFT and DFT is employed, and final predictions are obtained after passing through a fully connected layer. Evaluation is conducted on three datasets: the COVID-19 datasets of 38 African countries and 42 European countries from worldometers, and the chickenpox dataset of 20 Hungarian regions from Kaggle. Metrics include Average Root Mean Square Error (ARMSE) and Average Mean Absolute Error (AMAE).ResultFor African COVID-19 dataset and Hungarian Chickenpox dataset, GRGNN consistently outperforms other methods in ARMSE and AMAE across various prediction step lengths. Optimal results are achieved even at extended prediction steps, highlighting the model’s robustness.ConclusionGRGNN proves effective in predicting epidemic time series data with high accuracy, demonstrating its potential in epidemic surveillance and early warning applications. However, further discussions and studies are warranted to refine its application and judgment methods, emphasizing the ongoing need for exploration and research in this domain.

Outlier Detection for Permanent Magnet Synchronous Motor (PMSM) Fault Detection and Severity Estimation

Today, Permanent Magnet Synchronous Motors (PMSMs) are a dominant choice in industry applications. During operation, different possible faults in the system can occur, so early and automated fault detection and severity estimation are required to ensure smooth operation and optimal maintenance planning. In this direction, outlier detection methods are employed in this paper. The motor’s current signals are used to extract useful indicators of the fault, along with d-q transform. Statistical indicators in both time and frequency domains are selected to describe fault-related patterns. Based on the extracted features, three outlier detection methods are investigated: the Isolation Forest, the One Class Support Vector Machine, and the Robust Covariance Ellipse. Each method is investigated through different model parameters to evaluate fault detection and severity estimation capabilities. Finally, an ensemble approach is proposed based on decisions and outlier score ensemble. The proposed methodology is verified through different operating conditions in a PMSM test bench.

Fractional differentiation-based variational level set model for noisy image segmentation without contour initialization

Abstract Noisy image segmentation is a hot topic in image analysis. In this paper, we present a novel methodology for tackling this issue through the integration of fractional differentiation in the frequency domain with a variational level set model, which eliminates user-selected initial contours by incorporating the convex energy function. Additionally, the fractional differentiation reduces noises while preserving more detail information. Experiments on synthetic and real noisy images demonstrate that our proposed model surpasses other denoising variational level set models in terms of noise reduction, segmentation accuracy and efficiency.

Res2Net-based multi-scale and multi-attention model for traffic scene image classification.

With the increasing applications of traffic scene image classification in intelligent transportation systems, there is a growing demand for improved accuracy and robustness in this classification task. However, due to weather conditions, time, lighting variations, and annotation costs, traditional deep learning methods still have limitations in extracting complex traffic scene features and achieving higher recognition accuracy. The previous classification methods for traffic scene images had gaps in multi-scale feature extraction and the combination of frequency domain, spatial, and channel attention. To address these issues, this paper proposes a multi-scale and multi-attention model based on Res2Net. Our proposed framework introduces an Adaptive Feature Refinement Pyramid Module (AFRPM) to enhance multi-scale feature extraction, thus improving the accuracy of traffic scene image classification. Additionally, we integrate frequency domain and spatial-channel attention mechanisms to develop recognition capabilities for complex backgrounds, objects of different scales, and local details in traffic scene images. The paper conducts the task of classifying traffic scene images using the Traffic-Net dataset. The experimental results demonstrate that our model achieves an accuracy of 96.88% on this dataset, which is an improvement of approximately 2% compared to the baseline Res2Net network. Furthermore, we validate the effectiveness of the proposed modules through ablation experiments.