Increasing stability for inverse acoustic source problems in the time domain

DrowsyDG-Phys: Generalizable driver drowsiness estimation in conditional automated vehicles using physiological signals.

Parallel high-throughput system for time domain diffuse optical spectroscopy based on a 16-channel SiPM array

Epilepsy is a chronic neurological disease that profoundly impacts patients' daily lives. Electroencephalography (EEG) serves as a crucial tool for the clinical diagnosis of epilepsy and other brain disorders. Current research methods primarily concentrate on the time domain of EEG signals, often preprocessing frequency domain information without thorough exploration or effective integration with the time domain. To overcome the limitations of traditional models in extracting comprehensive frequency domain information and fusing time and frequency data, this paper proposes a Time-Frequency Cross-Attention Network (TFCANet) based on the residual attention mechanism. This network converts time-domain features into frequency-domain features using a Fast Fourier Transform. Subsequently, four SE Residual modules are employed to extract features for the frequency domain branch, while a Residual Window Multi-head Self-Attention (ResWMSA) mechanism is utilized for the time domain branch. Finally, cross-attention is applied to achieve inter-modal feature fusion. The proposed model is experimentally evaluated on the HMS-Harmful Brain Activity Classification dataset from Kaggle's 2024 competition and a dataset from the University of Bonn, Germany. Our model achieved 96.15% accuracy on a five-category task using the HMS dataset and 93.63% accuracy on a five-category task using the University of Bonn dataset. These results demonstrate that our model fully integrates features from both time and frequency domains, highlighting the superiority of time-frequency feature fusion over single-modality approaches for epilepsy detection.

Initial afferent activation of V1, indexed by the C1 component of the human VEP, is often considered to be a rudimentary stage of visual processing, operating mostly as a conduit for later stages with limited cognitive penetrability. The full suite of visual analysis entails activity across several visual areas and feedback from later areas to earlier ones. This raises the question of whether the early sensory representation indexed by the C1 is read out for perceptual decisions or whether it is passed over in favour of more advanced representations. To address this question, we asked whether the C1 would predict time-pressured stimulus contrast comparisons independently of physical stimulus conditions, a phenomenon known as choice probability. We found that the C1 did this for a narrow range of response times, indicative of decision readout since the C1 is a transient signal. This effect could not be accounted for by stimulus differences, choice history, or any other choice-predictive signal that we could identify in either the time or frequency domain, either before or after target onset. It also preceded the onset of evidence-dependent decision formation estimated from the centroparietal positivity by tens of milliseconds, together providing an approximate timeline of early evidence readout and its delayed impact on the decision.

Deep-rooted crops accessing water and nutrients from deep soil layers enhance the resource base for crop production. However, studying these roots in field conditions is labour-intensive, limiting research scope. We established a field root research facility with 48 plots for replicated experiments. The facility includes 144 6-metre-long minirhizotron tubes and an AI-based pipeline for rapid root trait analysis. We also attempted to install access tubes and customized ingrowth core production for less-invasive root activity determination. Our study revealed significant differences in deep root density among species, particularly at depths of 2.5 to 4.5 m, over 5 years. The less-invasive studies using ingrowth cores reached depths of 4.2 m. Nutrient tracer 15N analysis showed marked differences in deep root activity among crop species. Time domain reflectometry sensors indicated varying water depletion in deeper soil layers, influenced by crop species and root growth patterns. We established a field facility for studying deep root growth and function, demonstrating its effectiveness in analysing diverse deep-rooted plant species. This facility provides an ideal platform for conducting meaningful research in deep soil layers, yielding statistically and biologically significant results for agricultural applications.

Determination of torsion angles via recoupling of backbone HC and HN dipolar interactions is a well-known method in magic-angle spinning NMR spectroscopy. Torsion angle values can be obtained by comparing simulated and experimental signals, either in the frequency or time domains. Typically, all molecular orientations are assumed to have identical detected amplitudes at zero recoupling time. The changes in these amplitudes during the recoupling period define the dipolar coupling values and the torsion angles. Experimentally, however, orientations may exhibit different detected amplitudes due to additional cross-polarization (CP) blocks that connect different spins in multidimensional experiments. We numerically and experimentally investigate how CP blocks bias backbone φ torsion angle determination and propose CP conditions that minimize this effect, thereby improving accuracy. Applying these conditions in pseudo-4D (H)CANH experiments yields improved agreement of the extracted angles with X-ray crystallographic data for microcrystalline chicken α-spectrin SH3. For the influenza A M2 membrane protein, we identify an unexpected backbone dihedral angle for the I32 residue, which is consistent with TALOS-N predictions but deviates from ideal α-helical transmembrane geometry.

Abstract Reverberation mapping (RM) determines the mass of black holes (BHs) in active galactic nuclei (AGNs) by resolving the BH gravitational sphere of influence in the time domain. Recent RM campaigns have yielded direct BH masses through dynamical modeling for a sample of 32 objects, spanning a wide range of AGN luminosities and BH masses. In addition, accurate BH masses have been determined by spatially resolving the broad-line region with GRAVITY for a handful of AGNs. Here, we present a detailed analysis of Hubble Space Telescope images using surface-brightness profile fitting with state-of-the-art programs. We derive AGN luminosity and host-galaxy properties, such as radii and luminosities for the spheroid, disk, and bar (if present). The spheroid effective radii are used to measure stellar velocity dispersion from integral-field spectroscopy. Since the BH masses of our sample do not depend on any assumption of the virial factor needed in single-epoch spectroscopic mass estimates, we can show that the resulting scaling relations between the mass of the supermassive BHs and their host galaxies match those of quiescent galaxies, naturally extending to lower masses in these (predominantly) spiral galaxies. We find that the inner AGN orientation, as traced by the broad-line region inclination angle, is uncorrelated with the host-galaxy disk. Our sample has the most direct and accurate M BH measurements of any AGN sample and provides a fundamental local benchmark for studies of the evolution of massive BHs and their host galaxies across cosmic time.

Abstract Traditional water pipeline leakage identification has several limitations, such as incomplete signal feature extraction and the underutilization of physical information. To overcome these limitations, this study proposes Multimodal Two-stream Feature-fusion with Two-stage Convolutional Neural Network (MTFT-CNN). The analysis reveals that local dynamic features in time-frequency spectrograms complement global static features in the time and frequency domains. Accordingly, a multimodal two-stream architecture is developed to extract global statistical features concurrently from vibration–acoustic signals and local dynamic features from spectrograms. The cross-modal bidirectional attention module enables the adaptive fusion of these heterogeneous representations. The two-stage classification strategy enhances recognition efficiency: in the first stage, a hard-gating mechanism filters out non-leakage samples, while in the second stage, a multi-task learning framework, guided by the physical correlation between pressure and aperture, achieves the fine-grained identification of operating conditions. The experimental results show accuracies of 99.15%, 97.75% and 96.63% for leakage state, pressure and aperture recognition, respectively, with an overall accuracy of 98.37%. Ablation and comparative analyses confirm that MTFT-CNN significantly improves the comprehensiveness, precision and engineering applicability of leakage information identification, offering an effective solution for high-precision and high-efficiency leakage information identification in water supply pipelines.

Metal additive manufacturing (AM) facilitates the production of geometrically complex components, yet its broader industrial use remains limited by the risk of defect formation and uncertainties in their detection, originating from the highly dynamic and high-temperature process environment. To make additive manufacturing more reliable and establish high-quality parts, it is important to understand how these defects form and how their characteristics appear during the process. This review explains the main causes of common defects, such as cracking, porosity, lack of fusion, and inclusions in metal AM processes, including Powder Bed Fusion and Directed Energy Deposition. It also connects main defect formation mechanisms to the optical, thermal, acoustic, and spectroscopic signals that can be measured during the process. Moreover, it is described how commonly used in situ monitoring systems work and how their signals correspond to melt pool dynamics, vapor plume, particle movement, and the solidification process for each kind of defect. An overview is provided of how data from these systems are analyzed, including the extraction of features from images, the evaluation of temperature fields, and the use of time and frequency domain techniques for various signals. By linking the physics of defect formation to measurable process signals, the interpretation of sensor data is enabled, and potential strategies for monitoring specific problems are outlined. Finally, recent developments are examined, including the integration of multiple sensors, advanced feature-representation approaches, and real-time data interpretation coupled with adaptive control. Together, these directions represent promising advances towards more intelligent and reliable monitoring systems for the future of metal AM.

Identifying the time-domain waveform type under broadband conditions is a basic but very challenging task. Traditional methods based on frequency domain or training models generally have the problems of high resource consumption, large delay, and unsuitability for hardware. This paper proposes a time-domain waveform recognition architecture based on an FPGA, which is integrated with multi-feature voting. Several lightweight time domain characteristics, such as high amplitude ratio, symmetry, slope uniformity, slope change rate, and flat-top characteristics, are extracted and directly used for waveform classification. Then classify sine waves, square waves, triangular waves, and noise in the time domain according to the decision-making mechanism of voting. In order to improve reliability under non-ideal conditions, adaptive thresholds and noise perception decision-making logic are used to suppress misclassifications caused by random fluctuations and jitter. The whole engineering design focuses on resource consumption and hardware efficiency, using a fully pipeline FPGA architecture. The experimental results prove that the system has the ability of high-precision identification, low power consumption, and real-time processing in the wide frequency band, providing an efficient and practical solution for embedded waveform recognition applications.

Earthquake monitoring systems (EQS) play a critical role in seismic hazard assessment and tectonic studies. A complete and accurate earthquake catalog improves our understanding of fault mechanisms and supports the development of strategies to enhance public safety and the resilience of infrastructure. We propose a foundation model to improve EQS performance through a two-stage framework. Stage 1 involves training the foundation model in a self-supervised manner using a U-Trans architecture, which combines a U-Net encoder-decoder structure with a compact convolutional transformer in the bottleneck layer. At this stage, the model learns to reconstruct corrupted seismic waveforms in both the time and frequency domains, enabling it to extract well-structured and informative features in the latent space. Stage 2 focuses on downstream earthquake tasks. The latent features extracted by the encoder are flattened and concatenated as an additional input channel to the downstream models, effectively guiding them toward better performance. The U-Trans network is trained on more than 2 million three-component seismograms collected from three open-source datasets, ensuring diverse coverage and robust feature extraction. Extensive testing demonstrates that incorporating the encoder's latent features significantly improves the performance of several key downstream tasks, including seismic phase picking, earthquake location, magnitude estimation, and P-wave polarity classification. Analysis of the latent space reveals that the extracted features strongly correspond to P- and S-wave arrival times, which are crucial for many earthquake monitoring applications.

Abstract In this paper, a novel time domain sampling method based on the initial arrival time of waves is proposed to reconstruct acoustic sources, including point sources, curve sources, surface sources and block sources. The uniqueness of reconstructing sources whose spatial support is a convex region is proved. Theoretical analyses are provided to demonstrate the validity of the proposed sampling method in reconstructing various types of sound sources. The proposed algorithm does not involve the time integral, exhibits high computational efficiency, and demonstrates strong noise resistance. Numerical experiments are conducted to show the effectiveness of the proposed method.

Sequential recommendation systems have become essential for personalized services in e-commerce and content platforms. While recent research has extended these systems with multi-modal features, existing approaches face three major challenges. First, they inadequately model fine-grained temporal interval distributions, failing to discriminate between high-frequency short intervals and low-frequency long intervals. Second, uniform fusion in the time domain leads to semantic misalignment across modalities because it ignores their inherent differences in the frequency domain. Third, rigid fusion strategies without self-supervised constraints lead to limited representation quality and semantic drift from pre-trained embeddings. To address these issues, we propose ATHWE, an A daptive T emporal Expert Routing with H ierarchical W avelet E nhancement framework. ATHWE employs exponential saturation time mapping to generate temporally adaptive embeddings. These embeddings guide a sparse mixture of experts to model multi-scale user behavior dynamics. A hierarchical wavelet decomposition with band-specific gating selectively fuses complementary frequency components across modalities. Furthermore, contrastive learning and cluster-preserving objectives preserve semantic information during multi-modal fusion. Extensive experiments on multiple datasets validate the effectiveness of our framework. Our code is available at https://github.com/lulusiyuyu/ATHWE .

We propose a causality-based divide-and-conquer algorithm for nonequilibrium Green’s function calculations with quantics tensor trains. This algorithm enables stable and efficient extensions of the simulated time domain by exploiting the causality of Green’s functions. We apply this approach within the framework of nonequilibrium dynamical mean-field theory to the simulation of quench dynamics in symmetry-broken phases, where long-time simulations are often required to capture slow relaxation dynamics. We demonstrate that our algorithm allows to extend the simulated time domain without a significant increase in the cost of storing the Green’s function.

Compared with rule-based and optimization energy management strategies, online optimal energy management control strategies for a dual-mode power-split hybrid electric vehicles (PSHEVs) are able to achieve better fuel economy and real-time performance. Global online optimization of a finite time domain energy management strategy based on the hybrid model predictive control (HMPC) algorithm is proposed in this study. To reduce the computing time, a linearized predictive model is built; because dual-mode PSHEVs can be considered hybrid systems that include continuous and discrete states, the hybrid states can be expressed uniformly. Therefore, a mixed logical dynamic (MLD) predictive model is built based on hybrid system theory, and an HMPC energy management strategy is proposed based on the MLD predictive model. To solve the optimal control problem online to obtain the optimal control sequence, the optimal control problem is converted into a mixed-integer linear programming (MILP) problem. The HMPC-based energy management strategy is compared with dynamic programming (DP)-based and rule-based energy management strategies over two different driving cycles. Simulation results indicate that the HMPC-based EMS achieves 80.60% and 83.79% of the fuel economy performance obtained by the DP-based EMS. In comparison, the rule-based EMS only achieves 66.46% and 70.51% of the DP-based control performance. Therefore, the HMPC-based energy management strategy is favorable for real-time control while effectively improving fuel economy.

This article describes the theory of the time-based frequency-locked loop (TFLL), intended for tracking the input signal periods, which change extremely rapidly. They are based on the processing of the input signal periods. The described algorithm, when appropriately implemented, can also be applied to track and predict any physical variable in real time, provided that the pulse signal period is replaced with periodic data from the physical variable. The paper presents a mathematical procedure for determining optimal system parameters for tracking rapidly changing periods. Since this mathematical procedure is very lengthy for higher-order systems, the paper focuses on a new abbreviated analysis method based on a tabular approach. This approach has not been described in the literature so far. Mathematical analyses in the time domain were performed using the Z transformation. Simulation of the system operation was carried out. For frequency-domain analysis, the theory of FIR digital filters and the corresponding MATLAB software were used. The system's tracking capabilities are demonstrated in the time and frequency domains.

With the development of society, the incidence of cardiovascular diseases has continued to rise. As an important non-invasive diagnostic tool, electrocardiography (ECG) has been widely used in clinical screening and auxiliary diagnosis. In recent years, deep learning-based automated ECG analysis methods have attracted considerable attention; however, most existing approaches rely on single time-domain features, making it difficult to fully exploit the discriminative information of ECG signals in both time and frequency domains. Therefore, this study proposes a Frequence-Enhance ECG Network (FEENet), which performs multi-scale feature extraction on both temporal ECG signals and their frequency-domain representations. A Transformer-based causal encoder (TCE) is then employed to model the relationships among features at different scales. Subsequently, a time–frequency cross-attention (TFCA) module is introduced to enable bidirectional interaction and deep fusion between temporal and frequency-domain features. Finally, a classification layer is used to produce accurate cardiovascular disease classification results. The proposed method achieves an accuracy of 89.6% and an F1-score of 77.4% on the PTB-XL dataset, demonstrating its effectiveness and strong classification capability in complex ECG classification tasks.

To ascertain the cross-sectional association between cardiovascular autonomic dysfunction and arterial stiffness across glucose metabolism status. We performed a cross-sectional analysis of participants of the Maastricht study. Cardiovascular autonomic function was based on heart rate variability (HRV) indices from 24-hour ECG recordings and summarized in z-scores for time and frequency domains. Aortic and carotid stiffness were assessed by carotid-femoral pulse wave velocity (PWV) and carotid artery distensibility (CD), respectively. We used multiple linear regression to study the associations and adjusted for demographic and lifestyle factors and a range of cardiovascular risk factors. We tested for effect modification of the associations by glucose metabolism status. PWV and CD measures were available in 3673 and 1802 participants, respectively (median (25th; 75th percentile) age: 60 years (53; 66), 51% women, 20% type 2 diabetes by design. Participants with lower HRV had higher aortic stiffness. Per SD lower time-domain and frequency-domain HRV z-scores were associated with 2.8% (95% CI 2.1% to 3.4%) and 2.8% (95% CI 2.1% to 3.5%) higher PWV, respectively. Similar trends were observed for carotid stiffness, with 3.2% (95% CI 1.4% to 5.0%) and 3.1% (95% CI 1.2% to 5.0%) lower CD per SD lower time-domain and frequency-domain HRV, respectively. The magnitude of these associations was higher in groups with prediabetes and type 2 diabetes compared with those with normal glucose metabolism, with evidence of effect modification by glucose metabolism status (p value for interaction: <0.01 for prediabetes and <0.05 to <0.10 for type 2 diabetes, both compared with normal glucose metabolism). Cardiovascular autonomic dysfunction is associated with higher aortic and carotid stiffness, especially in people with dysglycemia.

Abstract As most of the companions in the double neutron star systems should be normal pulsars, the Fast Folding Algorithm (FFA), which is suitable for finding these long spin period pulsars, was used to search their possible radio signals. A time domain resampling code PYSOLATOR was used to maximize the available data length by removing the orbital modulation. We collected and processed 272.2 hours observational data taken by the Five-hundred meter Aperture Spherical radio Telescope (FAST) for the 13 double neutron star systems in its sky. The signal-to-noise ratios of known pulsar signals are obviously improved by this search method, including the detection of a faint pulsar signal which only saw by folding the data. Unfortunately, no companion signals were found among all the 197962 candidates. Geodetic precession of the orbit could enhance detectability in future observations.