Noise Detection Research Articles

High-robust compressive multimode fiber imaging based on observation vector feedback correction

Multimode fiber (MMF) imaging technology has promising applications in endoscopic systems, and the compressive multimode fiber imaging (CMMFI) scheme has garnered considerable attention due to its simple optical setup and fast imaging speed. However, the imaging quality of this scheme is significantly impacted by detector noise and environmental interference. This paper introduces a sampling method and reconstruction algorithm that utilizes edge detection and region growing techniques for feedback-based correction of the observation vector data acquired from the detector, aiming to reduce the interference of noise in the system. Simulation and experimental tests were conducted to validate this method, and the results underwent statistical analysis. The findings demonstrate that the proposed method improves imaging quality of the low-cost CMMFI without the need for increasing the number of samples. Moreover, it exhibits robustness against noise, further enhancing the reliability of the imaging process. Overall, this research represents a valuable contribution to the advancement of robust endoscopy based on multimode fibers.

Cross-sectional study on the health of workers exposed to occupational noise in China.

This study aimed to understand the health of workers exposed to occupational noise and explore the influencing factors related to workers' health, especially the impact of noise on workers' hearing. This work can provide a basis for formulating relevant measures for occupational noise prevention and control in the future. On the basis of the key occupational disease monitoring project in Chongqing, China, in 2021, the data of 1125 workers exposed to occupational noise were analyzed. Data included demographic information, occupational history, clinical physical examination information, and noise detection information of the working environment. Chi-square test and multifactorial logistic regression were used for statistical analysis. The prevalence rates of abnormal electrocardiogram (ECG), blood pressure (BP), and pure tone audiometry (PTA) were 21.9% (246/1125), 27.8% (313/1125), and 18.0% (202/1125), respectively. Male workers accounted for 78.8%. Compared with male workers, female workers had a lower prevalence of abnormal PTA (OR = 0.28, 95% CI = 0.16-0.50). Workers working in medium enterprises had a lower prevalence of abnormal BP than workers in micro enterprises (OR = 0.36, 95% CI = 0.19-0.66). The prevalence of abnormal BP and PTA of workers increased with age. After adjusting for age, sex, and body mass index, the prevalence of abnormal ECG of mining workers was higher than that of manufacturing workers (OR = 1.54, 95% CI = 1.07-2.24), and the prevalence of abnormal PTA had a rising trend with the increase in noise exposure value. Noise-exposed workers have a high prevalence of abnormal ECG, BP, and PTA, and factors such as age, enterprise size, and workplace noise exposure are correlated with the aberrant health of workers. Governments, enterprises, and individuals need to attach great importance to the possible adverse effects of noise. They must also actively adopt various effective measures to protect the occupational safety and health of workers.

Deep learning with noisy labels in medical prediction problems: a scoping review.

Medical research faces substantial challenges from noisy labels attributed to factors like inter-expert variability and machine-extracted labels. Despite this, the adoption of label noise management remains limited, and label noise is largely ignored. To this end, there is a critical need to conduct a scoping review focusing on the problem space. This scoping review aims to comprehensively review label noise management in deep learning-based medical prediction problems, which includes label noise detection, label noise handling, and evaluation. Research involving label uncertainty is also included. Our scoping review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. We searched 4 databases, including PubMed, IEEE Xplore, Google Scholar, and Semantic Scholar. Our search terms include "noisy label AND medical/healthcare/clinical," "uncertainty AND medical/healthcare/clinical," and "noise AND medical/healthcare/clinical." A total of 60 papers met inclusion criteria between 2016 and 2023. A series of practical questions in medical research are investigated. These include the sources of label noise, the impact of label noise, the detection of label noise, label noise handling techniques, and their evaluation. Categorization of both label noise detection methods and handling techniques are provided. From a methodological perspective, we observe that the medical community has been up to date with the broader deep-learning community, given that most techniques have been evaluated on medical data. We recommend considering label noise as a standard element in medical research, even if it is not dedicated to handling noisy labels. Initial experiments can start with easy-to-implement methods, such as noise-robust loss functions, weighting, and curriculum learning.

Speckle Noise Detection and Removing by Machine Learning Algorithms in Multisensory Images for 5G Transmission

The multispectral satellite sensor images have multibands, which have some typical noise. There is difficult to detect this tipical noise with low resolution image. The satellite local or gloval pixel information and quantificationare degraded due to this noise. Many standard transformations and filtering operations are developed for detection and removing of non-gaussion noise, which are not given sophisticated results with existing methods. These statistical characteristics are applied to those samples to identify and quantify present tipical noise. The higher-order statistical based machine learning algorithm is developing to remove the speckle noise from satellite image. In this proposed algorithm, implemented the higher order statistical approache such as skewness, kurtosis based adaptive curvelet methods are implemented for the detection and suppression of speckle noise with retrieve of spectral and spectral values. The proposed algorithm preserves smooth and sharp details and maintains the tradeoff level in multispectral bands is suitable for advanced high speed 5G communication with the effective rate of transmission. The proposed results are verified with suitable statistical parameters such as PSNR, Entropy and ERGAS values.

Speckle noise detection and correction for frequency-scanning interferometry in vibration measurement

Speckle noise significantly impairs the accuracy of vibration measurements for non-cooperative targets in frequency-scanning interferometry (FSI). To restore the real vibration, this paper proposes a comprehensive vibration demodulation algorithm based on the FSI. We established an FSI vibration model accounting for speckle noise. The extraction of instantaneous frequency from the interference signal is performed using a complex-shifted Morlet wavelet. An autoregressive differential model, incorporating a smooth weight function, is applied for bidirectional prediction to correct speckle noise. Finally, the true vibration of the target is restored through Kalman filtering. We validated the algorithm’s accuracy using a Monte Carlo strategy. The experimental results show that the surface vibration of black aluminum oxide alloy and the axial runout of the DC motor during operation was successfully captured, and the time–frequency analysis results showed that this method has strong robustness against dense speckle noise.

Identifying Plausible Labels from Noisy Training Data for a Land Use and Land Cover Classification Application in Amazônia Legal

Most studies in the field of land use and land cover (LULC) classification in remote sensing rely on supervised classification, which requires a substantial amount of accurate label data. However, reliable data are often not immediately available, and are obtained through time-consuming manual labor. One potential solution to this problem is the use of already available classification maps, which may not be the true ground truth and may contain noise from multiple possible sources. This is also true for the classification maps of the MapBiomas project, which provides land use and land cover (LULC) maps on a yearly basis, classifying the Amazon basin into more than 24 classes based on the Landsat data. In this study, we utilize the Sentinel-2 data with a higher spatial resolution in conjunction with the MapBiomas maps to evaluate a proposed noise removal method and to improve classification results. We introduce a novel noise detection method that relies on identifying anchor points in feature space through clustering with self-organizing maps (SOM). The pixel label is relabeled using nearest neighbor rules, or can be removed if it is unknown. A challenge in this approach is the quantification of noise in such a real-world dataset. To overcome this problem, highly reliable validation sets were manually created for quantitative performance assessment. The results demonstrate a significant increase in overall accuracy compared to MapBiomas labels, from 79.85% to 89.65%. Additionally, we trained the L2HNet using both MapBiomas labels and the filtered labels from our approach. The overall accuracy for this model reached 93.75% with the filtered labels, compared to the baseline of 74.31%. This highlights the significance of noise detection and filtering in remote sensing, and emphasizes the need for further research in this area.

Enhancing empirical software performance engineering research with kernel-level events: A comprehensive system tracing approach

Performance engineering is a proactive and systematic approach aimed at designing, building, and enhancing software systems to ensure their efficient and reliable operation. It involves observing and measuring the operational behavior of a software system without interference, assessing performance metrics like response times, throughput, and resource utilization. This entails delving into kernel-level events related to performance monitoring, which play a significant role in understanding system behavior and diagnosing performance-related issues. Kernel-level events offer insights into how both the operating system and hardware resources are utilized. This information empowers system administrators, developers, and performance analysts to optimize and troubleshoot the system effectively.A critical aspect of performance analysis is root cause analysis, which involves delving deep into kernel-level events connected to performance monitoring. These events provide valuable insights into the utilization of operating system and hardware resources, equipping system administrators, developers, and performance analysts with tools to effectively troubleshoot and optimize the system. Our study introduces an innovative artifact that captures kernel-level events using Elasticsearch and Kibana, facilitating comprehensive performance analysis under diverse scenarios. By defining both Light-load and Heavy-load scenarios and simulating CPU, I/O, Network, and Memory noise, we offer researchers a realistic environment to explore innovative approaches to system performance enhancement.The artifact comprises both kernel events and system calls, resulting in a cumulative count of 24,263,691 events. The proposed artifact can serve three distinct applications. The first application emphasizes performance analysis by utilizing kernel events for monitoring. The second application targets noise detection and root cause analysis, again using kernel events. Finally, the third application investigates software phase detection through monitoring at the kernel level. These applications demonstrate that through our artifact, researchers can effectively analyze performance, detect and address performance noise, and identify software phases, contributing to the advancement of performance engineering methodologies.All the system configurations, scripts, and traces can be found in the artifact GitHub repository.11URL: https://github.com/mnoferestibrocku/dataset-repo.

High-Precision Fiber Noise Detection and Comparison over a 260 km Field Fiber Link.

In this paper, we present a high-precision optical frequency noise detection and comparison technique using a two-way transfer method over a 260 km field fiber link. This method allows for the comparison of optical frequencies between remote optical references without the need for data transfer through communication. We extend a previously established two-way comparison technique to obtain all data at the local site. Two optical carrier signals are injected into the bidirectional fiber from both ends, and one carrier is reflected back from the remote end. This enables the phase comparison of the two carrier signals at a single site without the need to transmit experimental data. The common-mode frequency noise induced by the bidirectional fiber link is detected and effectively suppressed without the need for sophisticated active fiber noise control. Our demonstration system, which uses a 260 km field fiber link and a common laser source, achieves a fractional instability of 2.5×10-17 at 1 s averaging time and scales down to 3.5×10-21 at 8000 s. This scheme offers the distinct advantage of completing the comparison at a single site, eliminating the need for remote data transfer via communication. This method is expected to enhance reliability for high-precision frequency comparisons between remote optical clocks and advanced atomic clocks.

Closeby Habitable Exoplanet Survey (CHES). I. Astrometric Noise and Planetary Detection Efficiency Due to Stellar Spots and Faculae

The Closeby Habitable Exoplanet Survey (CHES) is dedicated to the astrometric exploration for habitable-zone Earth-like planets orbiting solar-type stars in close proximity, achieving unprecedented microarcsecond precision. Given the elevated precision, meticulous consideration of photocenter jitters induced by stellar activity becomes imperative. This study endeavors to model the stellar activity of solar-type stars, compute astrometric noise, and delineate the detection limits of habitable planets within the astrometric domain. Simulations were conducted for identified primary targets of CHES, involving the generation of simulated observed data for astrometry and photometry, accounting for the impact of stellar activity. Estimation of activity levels in our sample was achieved through chromospheric activity indices, revealing that over 90% of the stars exhibited photocenter jitters below 1 μas. Notably, certain proximate stars, such as α Cen A and B, displayed more discernible noise arising from stellar activity. Subsequent tests were performed to evaluate detection performance, unveiling that stellar activity tends to have a less pronounced impact on planetary detectability for the majority of the stars. Approximately 95% of the targets demonstrated a detection efficiency exceeding 80%. However, for several cold stars, e.g., HD 32450 and HD 21531, with the habitable zones close to the stars, a reduction in detection efficiency was observed. These findings offer invaluable insights into the intricate interplay between stellar activity and astrometric precision, significantly advancing our understanding in the search for habitable planets.

Robustness of Deep Learning models in electrocardiogram noise detection and classification

Background and Objective: Automatic electrocardiogram (ECG) signal analysis for heart disease detection has gained significant attention due to busy lifestyles. However, ECG signals are susceptible to noise, which adversely affects the performance of ECG signal analysers. Traditional blind filtering methods use predefined noise frequency and filter order, but they alter ECG biomarkers. Several Deep Learning-based ECG noise detection and classification methods exist, but no study compares recurrent neural network (RNN) and convolutional neural network (CNN) architectures and their complexity. Methods:This paper introduces a knowledge-based ECG filtering system using Deep Learning to classify ECG noise types and compare popular computer vision model architectures in a practical Internet of Medical Things (IoMT) framework. Experimental results demonstrate that the CNN-based ECG noise classifier outperforms the RNN-based model in terms of performance and training time. Results:The study shows that AlexNet, visual geometry group (VGG), and residual network (ResNet) achieved over 70% accuracy, specificity, sensitivity, and F1 score across six datasets. VGG and ResNet performances were comparable, but VGG was more complex than ResNet, with only a 4.57% less F1 score. Conclusions:This paper introduces a Deep Learning (DL) based ECG noise classifier for a knowledge-driven ECG filtering system, offering selective filtering to reduce signal distortion. Evaluation of various CNN and RNN-based models reveals VGG and Resnet outperform. Further, the VGG model is superior in terms of performance. But Resnet performs comparably to VGG with less model complexity.

Baseline correction for the infrared spectra of exhaled breath

Infrared spectroscopy appears to be a promising analytical method for the metabolic analysis of breath. However, due to the presence of trace amounts in exhaled breath, the absorption strength of the metabolites remains extremely low. In such low detection limits, the nonlinear detection sensitivity of the infrared detector and electronic noise strongly modify the baseline of the acquired infrared spectra of breath. Fitting the reference molecular spectra with the baseline-modified spectral features of breath metabolites does not provide accurate identification. Therefore, baseline correction of the acquired infrared spectra of breath is the primary requirement for the success of breath-based infrared diagnosis. A selective spectral region-based, simple baseline correction method is proposed for the infrared spectroscopy of breath.

Performance evaluation and investigation of diffraction optical elements effect on bit error rate of free space optics and performance investigation of space uplink wireless optical communication under varying atmospheric turbulence conditions

AbstractSeveral other optical antenna topologies have been developed and implemented throughout the years. These topologies include a variety of optical components, including the axicon optical element, dual‐secondary mirror, cone reflecting mirror, prism beam slier, and beam‐splitter/beam combiner. In contrast, the secondary reflecting mirror causes an obscuration loss that must be compensated for by reducing the transmission power in an optical antenna design. In order to address this issue in space optical communication, the present research helps to develop an enhanced two diffractive optical elements (DOEs) technology however the data presented therein only shows that DOEs may boost transmission power efficiency, which is insufficient for system designers. Though On‐Off Keying (OOK) is widely used in optical communication systems at the moment, the proposed research include DOEs into an OOK space uplink optical. The proposed research uses numerical simulation to explore how much a space uplink OOK system's bit error rate (BER) may be improved by using DOEs and adjusting fundamental parameters. The proposed BER model takes environmental factors like wind and detector noise into account. Using this theoretical model, the present work helps to investigate the effect of DOEs on the BER versus fundamental parameter characteristic curves in space uplink optical communication. Based on the findings, the DOEs structure has the potential to significantly enhance the BER performance of space uplink optical communication systems, especially at high obscuration ratios. When the obscuration ratio is 0.25, 0.167, or 0.125 and the transmission power is 1 W, for instance, the DOEs may improve the BER by a factor of two or one order of magnitude or less when the parameters are changed to the typical parameter values as specified. Results increase by a factor of six, three, and two orders of magnitude, respectively, when transmitting at 5 W. The results show that DOEs can significantly enhance the BER performance, especially at high obscuration ratios. The findings suggest that integrating DOEs into the optical subsystem is a straightforward approach to improving the performance of space uplink optical communication systems.

A novel framework and validation for the dynamic characterization of civil structures via ground-based lidar

While there has been substantial research conducted on the use of Ground-Based Lidar (GBL; i.e., Terrestrial Laser Scanners) in monitoring the static deformation of civil structures, its application in monitoring dynamic vibrations of structures, which is critical for structural health monitoring (SHM), has only been studied to a limited extent. Traditional contact-based SHM frameworks are constrained by a limited number of sensors and the need for physical access for instrumentation placement. Although recent case studies demonstrate GBL's potential in quantifying dynamic displacements, there is a need for more comprehensive research to validate the accuracy of GBL-based dynamic measurements under various GBL- and structure-based parameters in an autonomous and scalable framework. Therefore, the main objective of this study is to develop and comprehensively validate a novel end-to-end framework to monitor the dynamic vibrations of structures using GBL through extensive experimentation in a controlled laboratory environment. In this study, a novel two-step spatio-temporal algorithm was developed to extract the dynamic vibrations of structures from the dynamic point clouds. The framework leverages the Density-based Spatial Clustering of Applications with Noise (DBSCAN) and change detection algorithms. The impact of several GBL-based parameters on the accuracy of the operational modal analysis results was investigated across six single-degree-of-freedom structures with unique natural frequencies. The GBL-based parameters included the resolution, quality, and point-to-point distance of the dynamic point clouds. Accelerometers and infrared-based sensors were used for the validation of GBL measurements and operational modal analysis results. To validate GBL's full-field mode shapes across the tested specimen configurations, analytical and finite element models were constructed to provide high-fidelity mode shapes as ground-truth data. The results show that the GBL can detect sub-millimeter structural vibrations, and that the resulting natural frequencies and operational deflected shapes closely match those of traditional sensing modalities, and the analytical and finite element models. This study concludes that GBL can be used reliably for remotely monitoring the dynamic response of structures at a high spatial resolution. However, further research is warranted to evaluate the full extents of the proposed framework.

A stochastic gravitational wave background in LISA from unresolved white dwarf binaries in the Large Magellanic Cloud

ABSTRACT The Laser Interferometer Space Antenna (LISA) is expected to detect a wide variety of gravitational wave sources in the mHz band. Some of these signals will elude individual detection, instead contributing as confusion noise to one of several stochastic gravitational-wave backgrounds (SGWBs) – notably including the ‘Galactic foreground’, a loud signal resulting from the superposition of millions of unresolved double white dwarf binaries (DWDs) in the Milky Way. It is possible that similar, weaker SGWBs will be detectable from other DWD populations in the local Universe, including the Large Magellanic Cloud (LMC). We use the Bayesian LISA Inference Package (blip) to investigate the possibility of an anisotropic SGWB generated by unresolved DWDs in the LMC. To do so, we compute the LMC SGWB from a realistic DWD population generated via binary population synthesis, simulate 4 years of time-domain data with blip comprised of stochastic contributions from the LMC SGWB and the LISA detector noise, and analyse this data with blip’s spherical harmonic anisotropic SGWB search. We also consider the case of spectral separation from the Galactic foreground. We present the results of these analyses and show, for the first time, that the unresolved DWDs in the LMC will comprise a significant SGWB for LISA.

Top-Gated HfO₂-Capped InP Nanowire Photodetector With Low Noise and High Detectivity

Top-Gated HfO₂-Capped InP Nanowire Photodetector With Low Noise and High Detectivity

Neural axiom network for knowledge graph reasoning

Knowledge graph reasoning (KGR) aims to infer new knowledge or detect noises, which is essential for improving the quality of knowledge graphs. Recently, various KGR techniques, such as symbolic- and embedding-based methods, have been proposed and shown strong reasoning ability. Symbolic-based reasoning methods infer missing triples according to predefined rules or ontologies. Although rules and axioms have proven effective, it is difficult to obtain them. Embedding-based reasoning methods represent entities and relations as vectors, and complete KGs via vector computation. However, they mainly rely on structural information and ignore implicit axiom information not predefined in KGs but can be reflected in data. That is, each correct triple is also a logically consistent triple and satisfies all axioms. In this paper, we propose a novel NeuRal Axiom Network (NeuRAN) framework that combines explicit structural and implicit axiom information without introducing additional ontologies. Specifically, the framework consists of a KG embedding module that preserves the semantics of triples and five axiom modules that encode five kinds of implicit axioms. These axioms correspond to five typical object property expression axioms defined in OWL2, including ObjectPropertyDomain, ObjectPropertyRange, DisjointObjectProperties, IrreflexiveObjectProperty and AsymmetricObjectProperty. The KG embedding module and axiom modules compute the scores that the triple conforms to the semantics and the corresponding axioms, respectively. Compared with KG embedding models and CKRL, our method achieves comparable performance on noise detection and triple classification and achieves significant performance on link prediction. Compared with TransE and TransH, our method improves the link prediction performance on the Hits@1 metric by 22.0% and 20.8% on WN18RR-10% dataset, respectively.

Detection and characterization of ship underwater radiated narrowband noise

Underwater radiated noise (URN) emitted by ships is a high-intensity noise that interferes with acoustic transmissions. It is composed of broadband cavitation noise, tonal noise and its harmonics that can be observed up to several tens of kHz. The in-situ detection of ship noise and the identification of its tonal components are therefore useful for adaptive underwater communication techniques and for channel sensing. To this end, we present a detection scheme to identify the presence of a vessel and identifying its narrowband components. For presence detection, we develop a Convolution Neural Network (CNN) model whose input is a Detection Envelope Modulation On Noise (DEMON) analysis for a given observation window. The model is trained with our collected recordings of ship data and ambient noise. After identifying the ship’s URN, our tonal detector relies on the stability and stationarity of the ship’s tonal lines, as opposed to the randomness of the ambient noise. Cross-correlation and spectral entropy are used as detection metrics. To reduce the sensitivity to the tested environment, the detection thresholds are set adaptively. The results show a favorable trade-off between precision and recall compared to benchmark methods. We share our database of URN of labeled ships.

PSYCHOACOUSTICS-WEB: A free online tool for the estimation of auditory thresholds

PSYCHOACOUSTICS-WEB is an online tool written in JavaScript and PHP that enables the estimation of auditory sensory thresholds via adaptive threshold tracking. The toolbox implements the transformed up-down methods proposed by Levitt (Journal of the Acoustical Society of America, 49, 467-477, (1971) for a set of classic psychoacoustical tasks: frequency, intensity, and duration discrimination of pure tones; duration discrimination and gap detection of noise; and amplitude modulation detection with noise carriers. The toolbox can be used through a common web browser; it works with both fixed and mobile devices, and requires no programming skills. PSYCHOACOUSTICS-WEB is suitable for laboratory, classroom, and online testing and is designed for two main types of users: an occasional user and, above all, an experimenter using the toolbox for their own research. This latter user can create a personal account, customise existing experiments, and share them in the form of direct links to further users (e.g., the participants of a hypothetical experiment). Finally, because data storage is centralised, the toolbox offers the potential for creating a database of auditory skills.

Linear mathematical model for the underlying mechanism of extended ptychographic iterative engine

A linear model by writing the recorded diffractive intensities as a set of linear equations was put forwarded to determine the complex amplitude of the illumination and sample analytically for the extended ptychographic iterative engine (ePIE) imaging. An efficient computing method that can process large matrices was proposed to improve computing efficiency. The influences of detector noise and the positioning error of the sample were studied to explain why satisfactory reconstruction can be obtained experimentally when the linear equations of diffractive intensities are incorrect. This study demonstrates for the first time that ePIE technique has a unique and deterministic solution even with noisy experimental data, overcoming the main obstacle hindering the application of the ePIE algorithm in optical measurement and optical metrology, where a mathematically unique solution is crucial.

Direct Detection and Characterization of Exoplanets Using Imaging Fourier Transform Spectroscopy

Space-based direct imaging provides prospects for detection and spectral characterization of exoplanets at optical and near-infrared wavelengths. Integral field spectrographs (IFS) have been historically baselined for these mission concepts. However, multiple studies have revealed that detector noise is a serious obstacle for such instruments when observing extremely faint targets such as Earth-like planets. Imaging Fourier transform spectrographs (iFTS) are generally less sensitive to detector noise, and have several other compelling features such as simultaneous imaging and spectroscopy, smaller-format detector requirements, and variable spectral resolving power. To date, they have not been studied as options for such missions. Here, we compare the capabilities of IFS and iFTS for directly obtaining spectra from an Earth-like planet using both analytic and numerical models. Specifically, we compare the required exposure times to achieve the same signal-to-noise ratio with the two architectures over a range of detector and optical system parameters. We find that for a 6 m telescope, an IFS outperforms an iFTS at optical wavelengths due to the effects of distributed photon noise. In the near-IR, the relative efficiency of an IFS and iFTS depends critically on the instrument design and detector noise. An iFTS will be more efficient than an IFS if the readout noise of the near-IR detector is above ∼2–3 e− pix−1 frame−1 (t frm = 1000 s), which correspond to half to one-third of the state-of-art detector noise. However, if the readout noise is reduced below this threshold, the performance of an IFS will experience a substantial improvement and become more efficient. These results motivate consideration of an iFTS as an alternative option for future direct imaging space missions in the near-IR.