Clean Reference Research Articles

Deformation monitoring and data denoising without clean data for sandbars protective flexible mattress in the Yangtze River

The stability of the shoreline, such as sandbars, is vital for navigation and flood control in the Yangtze River, and the flexible mattresses are widely used to protect shorelines and submerged structures from erosion. To determine the long-term stability of the sandbars, it can be achieved by analyzing the monitoring data of the protective flexible mattress deformation. These mattresses are deployed in challenging environments, where traditional monitoring methods fall short in capturing their subtle, short-term deformations. To address this gap, this study introduces a novel deformation monitoring method based on Distributed Fiber Optic Sensing (DFOS) tailored for flexible mattresses and proposes a BiLSTM-CNN model for denoising the measured data, which circumvents the need for clean reference data. Key findings include: (1) For the monitoring of flexible mattresses, armored optical sensing cables should be installed from the perimeter inward, arranged in a U-shape post-submersion on the surface of the flexible mattress, and installed simultaneously with construction. The strain is measured by the single-end BOTDR. (2) The proposed BiLSTM-CNN model which is established without clean data, outperformed traditional denoising techniques like wavelet transform and Kalman filtering, effectively preserving essential strain responses while suppressing noise, enabling accurate depiction of the deformation of the flexible mattress, and demonstrating its applicability in complex monitoring scenarios. (3) The flexible mattress monitoring method was successfully deployed in the Yudaizhou shoreline protection engineering project. The BiLSTM-CNN denoising model, combined with a post-adjust method to process the different periods’ data, resulted in analyzable strain data. Comprehensive analysis of long-term strain monitoring data indicates that the flexible mattress in the monitoring area remains stable. This study’s significant contribution lies in advancing the methodology for monitoring and denoising strain data from underwater flexible mattresses, thereby enhancing the understanding and management of erosion protection systems in challenging environments.

Legacy effect of 25 years reduced atmospheric sulphur deposition on spruce tree nutrition

AbstractBackgroundSince the mid‐1990s, sulphur (S) pollution was drastically reduced in Central Europe. Over time, this has led to a distinct reduction in S availability for Norway spruce (Picea abies Karst.), which is still the most important timber species in Central European forestries.AimsDetermination of the Norway spruce nutritional status of former strongly affected areas by S pollution (Saxony) with different degrees of liming by assessing their foliar element contents and comparing them to regions remote from historical high S deposition.MethodsSites were selected based on levels of S deposition in the 1970–1990s with historical high deposition in Saxony (NE Germany), low deposition at Schluchsee (SW Germany) and Davos (Switzerland) as a clean air reference. Needles were sampled in late autumn 2019/2020 and elemental contents determined. Additional historical data on foliar S contents were available.ResultsHistorical data showed a clear decrease in foliar S contents in the Saxonian sites over the last 25 years, independent of liming. No difference between all study sites was found in the most recent sampling, whereas S together with other macronutrients strongly indicates deficiencies for forest growth and health.ConclusionsAfter 25 years of reduced S deposition, S nutrition became low for Norway spruce trees in Saxony, whereas soil parent material determines the overall tree nutritional status with respect to other nutrients. As such, no difference between sites with historical high, low or no S deposition was found. Further studies should focus on the mineralization of organic S in the topsoil to understand if S is effectively recycled within the forest ecosystem and on the effect of other diminishing nutrients such as Mg and P.

Utilizing the Concept of Entropy in the Second Law of Thermodynamics to Determine Risk Levels Associated with Groundwater and Bottled Drinking Water Using Rainwater Quality as a Clean Reference

The concept of entropy as a state function is employed in this paper within a radiation system designed to address the role of atomic disturbance in determining the levels of physical permeability changes in the studied states, comparing the results with those of a clean phase. This investigation begins with samples prepared for characterization tests using a unique technique for electromagnetic irradiation. The samples included: (1) a rainwater sample carefully collected during the first week of February 2024, serving as a high-quality natural reference in its aqueous phase; (2) a sample from a 42-meter-deep drinking water well (GW) in Ghout al-Dees, west of Tripoli, Libya; and (3) a bottled water sample (AA-WS3) commonly available in the Libyan market. The test results focused on the values of the physical property (τC), identified from the relationship between the entropy operator and time, as a physical fingerprint that accurately detects disturbed material permeability. The study concludes with a section linking the properties of the natural reference phase to the fingerprint characteristics of the studied samples, providing an approach to assess the quality of the tested water samples against the ideal chemophysical quality of rainwater. This led to findings that (1) the groundwater quality (QAA-WS2) was 76.92%, and (2) the quality of bottled water available in the Libyan market (QAA-WS3) was 74.57%, highlighting the role of the polymer composition in plastic container packaging as an additional factor, alongside the structures enclosing water sources, as in the case of the groundwater well exposed to contamination. Keywords: atomic disturbance, entropy, chemical identity, atomic irradiation, wave gaps

Recommendations for Using the Method of Standards in Bioindication Studies of the Northern Caspian Sea

Despite the increasing water and sediment pollution in the western part of the Northern Caspian Sea, the quality of its marine environment has never been assessed through bioindication, largely due to methodological issues. In this article, the method of standards, which involves comparing the species composition between the polluted and clean (reference) sites, was used for the bioindication of organic pollutants in the Northern Caspian Sea. The study area was segmented based on isobaths. The segments were determined depending on the range of depths inhabited by zoobenthos species and their ecological valence to oxygen. The depths with a sufficient abundance of oxyphilic zooplankton species (bioindicators) to apply the method of standards were located. The obtained results show that the method of standards is a good and reliable choice to assess the marine environment quality in the Northern Caspian Sea at isobaths of 6–9 m and within depths of 14–18 m for polluted and reference sites. However, the sediments must contain shells, sand, and siltstones. Therefore, while the method of standards has clear merits, it does not suffice for the rest of the Caspian Sea area.

Unsupervised speckle noise reduction technique for clinical ultrasound imaging.

Deep learning-based image enhancement has significant potential in the field of ultrasound image processing, as it can accurately model complicated nonlinear artifacts and noise, such as ultrasonic speckle patterns. However, training deep learning networks to acquire reference images that are clean and free of noise presents significant challenges. This study introduces an unsupervised deep learning framework, termed speckle-to-speckle (S2S), designed for speckle and noise suppression. This framework can complete its training without the need for clean (speckle-free) reference images. The proposed network leverages statistical reasoning for the mutual training of two in vivo images, each with distinct speckle patterns and noise. It then infers speckle- and noise-free images without needing clean reference images. This approach significantly reduces the time, cost, and effort experts need to invest in annotating reference images manually. The experimental results demonstrated that the proposed approach outperformed existing techniques in terms of the signal-to-noise ratio, contrast-to-noise ratio, structural similarity index, edge preservation index, and processing time (up to 86 times faster). It also performed excellently on images obtained from ultrasound scanners other than the ones used in this work. S2S demonstrates the potential of employing an unsupervised learning-based technique in medical imaging applications, where acquiring a ground truth reference is challenging.

Unsupervised denoising of the non-invasive fetal electrocardiogram with sparse domain Kalman filtering and vectorcardiographic loop alignment

Objective. Even though the electrocardiogram (ECG) has potential to be used as a monitoring or diagnostic tool for fetuses, the use of non-invasive fetal ECG is complicated by relatively high amounts of noise and fetal movement during the measurement. Moreover, machine learning-based solutions to this problem struggle with the lack of clean reference data, which is difficult to obtain. To solve these problems, this work aims to incorporate fetal rotation correction with ECG denoising into a single unsupervised end-to-end trainable method. Approach. This method uses the vectorcardiogram (VCG), a three-dimensional representation of the ECG, as an input and extends the previously introduced Kalman-LISTA method with a Kalman filter for the estimation of fetal rotation, applying denoising to the rotation-corrected VCG. Main results. The resulting method was shown to outperform denoising auto-encoders by more than 3 dB while achieving a rotation tracking error of less than 33∘. Furthermore, the method was shown to be robust to a difference in signal to noise ratio between electrocardiographic leads and different rotational velocities. Significance. This work presents a novel method for the denoising of non-invasive abdominal fetal ECG, which may be trained unsupervised and simultaneously incorporates fetal rotation correction. This method might prove clinically valuable due the denoised fetal ECG, but also due to the method’s objective measure for fetal rotation, which in turn might have potential for early detection of fetal complications.

Improved Reference Current Extraction Method for Electric Aircraft Active Power Filter Application

This paper proposes a robust reference current extraction method useful for active power filters (APFs) connected to the More Electric Aircraft (MEA) grid for addressing power quality issues. Hence, it is essential to investigate and develop better APF harmonic compensating control algorithms that can eliminate the impact of harmonics emerging from the nonlinear loads connected to the MEA grid. The proposed method envisages a half-cycle moving window discrete Fourier transform filter approach to rapidly extract the fundamental positive sequence (FPS) from a polluted three-phase grid voltage. Thus, allowing the harmonic cancelation method to generate desired clean reference current which is in phase with FPS component of the grid voltage. This reference current is used in conjunction with the hysteresis current controller to generate appropriate gate pulses for the current-controlled voltage source APF. The employment of the digital filter in the control structure of the APF ensures improved power quality for the MEA grid as compared to traditional filtering algorithms. The numerical and experimental results demonstrate that the proposed control solution helps to reduce the total harmonic distortion under abnormal grid disturbances, thus, indicating the potential benefits of the proposed work for addressing power quality issues in MEA applications.

Blockchain Based Data Steganography

Blockchain based Data Blockchain based Steganography is the system of hiding records which may be text, photo or video Interior a cowl photo. The secret facts are hidden in a way that it not seen to the human eyes. Deep gaining knowledge of era, which has emerged as a powerful tool in diverse programs along with image Blockchain based Steganography, has obtained increased interest lately. The principal purpose of this paper is to explore and talk numerous deep getting to know strategies available in image Blockchain based Steganography area. Deep gaining knowledge of techniques used for photograph Blockchain based Steganography can be broadly divided into 3 classes - traditional strategies, Convolutional Neural community-primarily based and general adverse community-based methods. Along with the technique, A problematic precis on the datasets used, experimentalu. S.A. Taken into consideration and the assessment metrics Normally used are defined in this paper. A table summarizing all of the info also are provided for clean Reference. This paper targets to assist the fellow researchers by means of compiling the current tendencies, challenges and few destinies course on this field

Analysis of PCDD/Fs in environmental samples by using gas chromatography in combination with high resolution mass spectrometry: optimization of sample preparation

ABSTRACT This study developed a simple, economical and time-saving clean-up procedure based on an in-house packed column system for the determination of the most toxic-relevant 17 PCDD (polychlorinated dibenzo-p-dioxin) and PCDF (polychlorinated dibenzofuran) congeners in environmental samples following accelerated solvent extraction. The extract solution was cleaned up using a manually packed column system consisting of a multi-layer silica gel column and an activated carbon column. After clean-up, dioxins and furans were analysed by using gas chromatography in combination with high-resolution mass spectrometry. Isotopic labelled spiking experiments investigated the recovery of all dioxin and furan congeners. Then, the clean-up procedure for environmental samples was validated by applying the analysis of both clean soil reference material (CRM EDF-5183) and heavily contaminated sediment reference material (CRM EDF-5184) samples, as well as participating in the proficiency test (PT) for sediment samples. The recovery efficiency of isotope-labelled standard compounds in certified reference material (CRM) samples and PT samples ranged from 35% to 101% and between 46.1% and 105%, respectively. The mass concentrations of 17 PCDD/Fs in CRM samples were in the acceptable range. The absolute z-score value for each PCDD/F congener in the PT samples was below 1, confirming the analytical procedure’s high accuracy. The applied clean-up procedure in this study has demonstrated accuracy, repeatability and a significant increase in sample extraction/preparation productivity by reducing the time for sample preparation.

Reconstructing echoes completely submerged in background noise by a stacked denoising autoencoder method for low-power EMAT testing

Low-power electromagnetic-acoustic transducer (EMAT) is crucially important for safety-critical equipment in industry, especially for potential explosives and inflammable petrochemical equipment and facilities. When the excitation power is very low, the corresponding echoes are overwhelmed in noise and related measurement would be inaccurate. To solve this problem, this paper presents a new echo reconstruction method based on a deep stacked denoising autoencoder (DSDAE) for nondestructive evaluation. First, the uses of reference signals and new data structure are to improve the training efficiency. A hybrid method based on variational mode decomposition and wavelet transform is used to obtain clean reference signals as inputs of the deep network. Then, the modified network structure and loss function aim to improve the ability of feature extraction and reconstruct clean echoes from low-power EMAT signals. To validate the effectiveness of the proposed method, the experiments of self-excitation and receiving A-scan inspections of stepped specimens with different thicknesses are conducted at some excitation voltages, as low as 25 V. The results indicate that the proposed DSDAE shows better and more stable denoising performance than some popular processing methods for different specimens and excitation voltages. It greatly improves the signal-to-noise ratio of the reconstructed signal to 20 dB. When applying to thickness measurement of specimens, its relative error is lower than 0.3%, which provides a practical and accurate tool for low-power EMAT testing.

A Self-Supervised Learning-Based Channel Estimation for IRS-Aided Communication Without Ground Truth

Deep learning (DL) is an emerging paradigm for accurate channel estimation for intelligent reflecting surface (IRS)-aided wireless communication systems. It has been proven to be a promising way to achieve better channel estimation performance for the IRS-aided wireless communication system than traditional methods (e.g., least-square algorithm). However, existing DL-based methods rely on ground truth (labels of the true channels) which is difficult to obtain in real networks. In this paper, we propose a self-supervised learning (SSL) method for the IRS channel estimation problem. No ground truth channel is needed in the training, while a simple and novel self-supervised denoising formula without a clean reference signal is presented. Particularly, in the training phase, the self-supervised signal and the input are the received signal vector and its noisy version, respectively. While in the inference phase the input is the estimated channel by using the least-square method and the output is the refined channel estimation. That is, our neural network-based channel estimation algorithm is not reciprocal for training and testing. We demonstrate that the proposed SSL solution has good convergence performance and generalization ability through numerical simulations. Interestingly, we find a “double descent” phenomenon in the learning curve during the test phase, i.e., when we gradually increase the number of training epochs, the performance first gets better, then becomes worse, and further gets better again. Besides, we propose to analyze SSL using the loss landscape and centered kernel alignment method. The results show that the self-supervised model has a similar loss landscape and representational similarity to the supervised model. We explored the effects of different signal-to-noise ratios (SNRs), different neural network sizes, and different training data volumes on our algorithm through numerical simulations. Extensive numerical simulation results show that our SSL algorithm is still competitive without ground truth. We also show that the developed scheme exhibits robustness to SNR ratio mismatch.

An automated workflow to assess completeness and curate GenBank for environmental DNA metabarcoding: The marine fish assemblage as case study

AbstractTo successfully implement environmental DNA‐based (eDNA) diversity monitoring, the completeness and accuracy of reference databases used for taxonomic assignment of eDNA sequences are among the challenges to be tackled. Here, we have developed a workflow that evaluates the current status of GenBank for marine fishes. For a given combination of species and barcodes, a gap analysis is performed and potentially erroneous sequences are identified. Our gap analysis based on the four most used genes (cytochrome c oxidase subunit 1, 12S rRNA, 16S rRNA, and cytochrome b) for fish eDNA metabarcoding found that COI, the universal choice for metazoans, is the gene covering the highest number of Northeast Atlantic marine fishes (70%), while 12S rRNA, the preferred region for fish‐targeting studies, only covers about 50% of the species. The presence of too close and too distant barcode sequences as expected by their taxonomic classification confirms the existence of erroneous sequences in GenBank that our workflow can detect and eliminate. Comparing taxonomic assignments of real marine eDNA samples with raw and clean reference databases for the most used 12S rRNA barcodes (teleo and MiFish), we confirmed that both barcodes perform differently and demonstrated that the application of the database cleaning workflow can result in drastic changes in community composition. Besides providing a tool for reference database curation, this study confirms the need to increase 12S rRNA reference sequences for European marine fishes and evidences the dangers of taxonomic assignments by directly querying GenBank. We have developed a workflow that evaluates the current status of GenBank for marine fishes. For a given combination of species and barcodes, a gap analysis is performed and potentially erroneous sequences are identified.

Non-intrusive speech quality assessment with attention-based ResNet-BiLSTM.

Speech quality is frequently affected by a variety factors in online conferencing applications, such as background noise, reverberation, packet loss and network jitter. In real scenarios, it is impossible to obtain a clean reference signal for evaluating the quality of the conferencing speech. Therefore, an effective non-intrusive speech quality assessment (NISQA) method is necessary. In this paper, we propose a new network framework for NISQA based on ResNet and BiLSTM. ResNet is utilized to extract local features, while BiLSTM is used to integrate representative features with long-term time dependencies and sequential characteristics. Considering that ResNet may result in the loss of context information when applied to the NISQA task, we propose a variant of ResNet which can preserve the time series information of the conferencing speech. The experimental results demonstrate that the proposed method has a high correlation with the mean opinion score of clean, noisy and processed speech.

Oxidative Stress in a Wild Population of the Freshwater Fish Hyphessobrycon Luetkenii Chronically Exposed to a Copper Mining Area: New Insights into Copper Toxicology.

In this study, we examine markers of oxidative stress in the tetra Hyphessobrycon luetkenii collected from two locations in the copper contaminated João Dias creek (southern Brazil). Also, specimens were translocated from a clean reference section of the creek to a polluted stretch and vice-versa. Fish were held at in submerged cages for 96h and then sacrificed. Nuclear abnormalities in erythrocytes and total antioxidant capacity, lipid peroxidation and protein carbonylation in gills, brain, liver and muscle displayed similar trends in both groups. Lipid peroxidation increased in all tissues of individuals translocated to the polluted site but only in liver and muscle of those translocated to the reference site. Increased protein carbonylation was also observed in gills of individuals translocated to the reference location. These results suggest similar oxidative stress among fish from the reference and polluted locations and that long-term metals exposure may require adaptations toward oxidative stress responses.

Using a single-channel reference with the MBSTOI binaural intelligibility metric

In order to assess the intelligibility of a target signal in a noisy environment, intrusive speech intelligibility metrics are typically used. They require a clean reference signal to be available which can be difficult to obtain especially for binaural metrics like the modified binaural short time objective intelligibility metric (MBSTOI). We here present a hybrid version of MBSTOI that incorporates a deep learning stage that allows the metric to be computed with only a single-channel clean reference signal. The models presented are trained on simulated data containing target speech, localised noise, diffuse noise, and reverberation. The hybrid output metrics are then compared directly to MBSTOI to assess performances. Results show the performance of our single channel reference vs MBSTOI. The outcome of this work offers a fast and flexible way to generate audio data for machine learning (ML) and highlights the potential for low level implementation of ML into existing tools.

Improved Image Quality Assessment by Utilizing Pre-Trained Architecture Features with Unified Learning Mechanism

The purpose of the no-reference image quality assessment (NR-IQA) is to measure perceived image quality based on subjective judgments; however, due to the lack of a clean reference image, this is a complicated and unresolved challenge. Massive new IQA datasets have facilitated the creation of deep learning-based image quality measurements. We present a unique model to handle the NR-IQA challenge in this research by employing a hybrid strategy that leverages from pre-trained CNN model and the unified learning mechanism that extracts both local and non-local characteristics from the input patch. The deep analysis of the proposed framework shows that the model uses features and a mechanism that improves the monotonicity relationship between objective and subjective ratings. The intermediary goal was mapped to a quality score using a regression architecture. To extract various feature maps, a deep architecture with an adaptive receptive field was used. Analyses of this biggest NR-IQA benchmark datasets demonstrate that the suggested technique outperforms current state-of-the-art NR-IQA measures.

Highly Stable Vibration Measurements by Common-path off-axis Digital Holography

Measurements of vibration and dynamic deformation with high resolution and accuracy by non-destructive techniques are of great interest in many branches of engineering. We report a non-destructive method based on digital holographic interferometry for high-precision micro-vibration measurement of a 3D object. Herein, a new and simple configuration of a common-path off-axis digital holographic system, with high mechanical stability and compact in size, is proposed. The simplicity, compactness, and high stability of the proposed setup are provided by employing a wedge plate in-line with the test object to create a clean reference beam. The object and reference beams propagate the same path, without the need for any specialized optical component or arrangement, and these beams allow to interfere on the faceplate of the image sensor to form a digital hologram. The feasibility of the proposed system is demonstrated by measuring the vibration of the objects driven by different excitation signals. A series of digital holograms of a vibrating object is recorded by the use of a high-speed image sensor. The result of the reconstruction of the recorded holograms is an array of complex numbers containing the complete amplitude and phase profile of the object, which in turn provides displacement and vibration information. Due to its simple geometry and significant outcomes, the proposed system could be established as a powerful tool for a wide range of optical imaging and measurement investigations.

TREE RINGS AS ARCHIVES OF ATMOSPHERIC POLLUTION BY FOSSIL CARBON DIOXIDE IN BRATISLAVA

ABSTRACTResults of radiocarbon (14C) analysis of a tree-ring series from Bratislava, Slovakia, covering the period from 1970 to 2004 are presented. For a part of this time period, monthly 14C measurements of atmospheric carbon dioxide from Bratislava sampling station are compared with the tree-ring results. The effects of fossil CO2 emissions on 14C levels in the environment are emphasized by comparison with atmospheric clean air reference values (Schauinsland, Germany). The presented results from Bratislava are also set against the previously measured tree-ring series from Low Tatras mountain region in Slovakia, representing regional clean air radiocarbon levels in the biosphere. The observed 14C levels of Bratislava tree rings and atmospheric CO2 in 1970s and 1980s are significantly lower than clean air data, indicating severe fossil CO2 pollution in Bratislava during this time period.

Non-intrusive deep learning-based computational speech metrics with high-accuracy across a wide range of acoustic scenes.

Speech with high sound quality and little noise is central to many of our communication tools, including calls, video conferencing and hearing aids. While human ratings provide the best measure of sound quality, they are costly and time-intensive to gather, thus computational metrics are typically used instead. Here we present a non-intrusive, deep learning-based metric that takes only a sound sample as an input and returns ratings in three categories: overall quality, noise, and sound quality. This metric is available via a web API and is composed of a deep neural network ensemble with 5 networks that use either ResNet-26 architectures with STFT inputs or fully-connected networks with wav2vec features as inputs. The networks are trained and tested on over 1 million crowd-sourced human sound ratings across the three categories. Correlations of our metric with human ratings exceed or match other state-of-the-art metrics on 51 out of 56 benchmark scenes, while not requiring clean speech reference samples as opposed to metrics that are performing well on the other 5 scenes. The benchmark scenes represent a wide variety of acoustic environments and a large selection of post-processing methods that include classical methods (e.g. Wiener-filtering) and newer deep-learning methods.

Method to detect carbon in silicon crystals in the concentration range down to 5 × 1014 cm−3 by Fourier transform infrared absorption at room temperature

The procedure of Fourier transform infrared absorption (FT-IR) at room temperature for quantifying C impurities in Si crystals was reexamined to improve the detection limit down to 5 × 1014 cm−3. Since the substitutional C absorption peak overlaps with a strong two-phonon absorption, the difference spectroscopy is necessary with using a C-lean reference sample. A baseline flatness of less than 0.0005 in absorbance and a thickness uniformity of less than 0.001 mm are required to realize a detection limit of 5 × 1014 cm−3 for 2 mm thick samples. To check the flatness, we define the Si/Si baseline which is the difference in the absorbance spectra measured twice with the removal and attachment of the same Si sample. Four organizations participated in the FT-IR round-robin test for ten samples with the C concentration ranging from 3.6 × 1014 to 3.3 × 1015 cm−3. The obtained C concentrations were almost within 30% deviation from the values determined by reliable secondary ion mass spectroscopy.