Classification on Underwater Acoustic Propagation Model U sing Convolutional Neural Network

Various line array configurations are evaluated for the source localization performance based on the analysis of mode decomposition matrix in this paper. The guideline of array shape design focuses on improving the localization performance of matched filed processing, meanwhile reducing the difficulty of deploying equipment in practical experiments. In the shallow water environment, when the environment is well known, the source localization result can be obtained by matched field processing algorithms effectively, but the source localization performance is affected by the array parameters, such as array length, the number of sensors, and the configurations of various horizontal and vertical line arrays. The modal decomposition method provides a useful insight into the questions of how many modes are needed and how to design the array to resolve the modes. Therefore, the method of utilizing a normal mode acoustic propagation model to decompose mode is proposed by vertical line array, horizontal line array and combined array respectively. Then we can evaluate the source localization performance of various line array configurations by studying the characteristic of normal mode decomposition matrix, thus establishing a qualitative or even quantitative relationship between each other. The more the normal mode decomposition matrix tends to be diagonalized, the better performance of line array localization will be obtained. Simulation results show that the localization performance of matched field processing with the combined arrays will be severely degraded when the mode amplitudes cannot be accurately deduced by one of the sub-arrays. Considering the requirements for the practical experiments and various environments, the source localization performance of short vertical line array and combined array are mainly discussed in this paper. The combined array can increase the azimuth and depth information of the source and realize three-dimensional target detection while the vertical array provides range-depth information and the horizontal array provides bearing information. Simulation result indicates that the design guidelines based on the normal mode decomposition are appropriate for arrays employed for matched filed processing. Meanwhile, the combined arrays perform better than the short vertical array, which is benefited by the horizontal array's suppressing the side lobes, which leads the ratio of peak to sidelobe to increase, and thus improving the location accuracy. The values of localization accuracy of combined arrays are all above 90% according to the simulation experiment. Take the practical application into account, the combined array is undoubtedly a compromise choice for the localization performance and the test complexity.

The papers in this special issue focus on the rapidly evolving techniques on shallow water acoustic communications, and how underwater acoustic (UWA) propagation models based on shallow water physics may be harnessed with signal processing techniques for high-resolution channel estimation, equalization, and resolving cross-layer challenges to achieve high-data-rate communications in shallow water. The definition of shallow water is relatively broad in scope, and not limited to any particular frequency range. Any UWA channel that exhibits significant multipath and rapid fluctuations in delay spread can be included in the paradigm of interest.

BackgroundThe aging of the world population poses a major health challenge, and brain–computer interface (BCI) technology has the potential to provide assistance and rehabilitation for the elderly.ObjectivesThis study aimed to investigate the electroencephalogram (EEG) characteristics during motor imagery by comparing young and elderly, and study Convolutional Neural Networks (CNNs) classification for the elderly population in terms of fatigue analysis in both frontal and parietal regions.MethodsA total of 20 healthy individuals participated in the study, including 10 young and 10 older adults. All participants completed the left- and right-hand motor imagery experiment. The energy changes in the motor imagery process were analyzed using time–frequency graphs and quantified event-related desynchronization (ERD) values. The fatigue level of the motor imagery was assessed by two indicators: (θ + α)/β and θ/β, and fatigue-sensitive channels were distinguished from the parietal region of the brain. Then, rhythm entropy was introduced to analyze the complexity of the cognitive activity. The phase-lock values related to the parietal and frontal lobes were calculated, and their temporal synchronization was discussed. Finally, the motor imagery EEG data was classified by CNNs, and the accuracy was discussed based on the analysis results.ResultFor the young and elderly, ERD was observed in C3 and C4 channels, and their fatigue-sensitive channels in the parietal region were slightly different. During the experiment, the rhythm entropy of the frontal lobe showed a decreasing trend with time for most of the young subjects, while there was an increasing trend for most of the older ones. Using the CNN classification method, the elderly achieved around 70% of the average classification accuracy, which is almost the same for the young adults.ConclusionCompared with the young adults, the elderly are less affected by the level of cognitive fatigue during motor imagery, but the classification accuracy of motor imagery data in the elderly may be slightly lower than that in young persons. At the same time, the deep learning method also provides a potentially feasible option for the application of motor-imagery BCI (MI-BCI) in the elderly by considering the ERD and fatigue phenomenon together.

This article present results from automatic detection of dust impact signals observed by the Solar Orbiter – Radio and Plasma Waves instrument. A sharp and characteristic electric field signal is observed by the Radio and Plasma Waves instrument when a dust particle impact the spacecraft at high velocity. In this way, ∼5–20 dust impacts are daily detected as the Solar Orbiter travels through the interstellar medium. The dust distribution in the inner solar system is largely uncharted and statistical studies of the detected dust impacts will enhance our understanding of the role of dust in the solar system. It is however challenging to automatically detect and separate dust signals from the plural of other signal shapes for two main reasons. Firstly, since the spacecraft charging causes variable shapes of the impact signals and secondly because electromagnetic waves (such as solitary waves) may induce resembling electric field signals. In this article, we propose a novel machine learning-based framework for detection of dust impacts. We consider two different supervised machine learning approaches: the support vector machine classifier and the convolutional neural network classifier. Furthermore, we compare the performance of the machine learning classifiers to the currently used on-board classification algorithm and analyze one and a half year of Radio and Plasma Waves instrument data. Overall, we conclude that classification of dust impact signals is a suitable task for supervised machine learning techniques. In particular, the convolutional neural network achieves a 96 % ± 1 % overall classification accuracy and 94 % ± 2 % dust detection precision, a significant improvement to the currently used on-board classifier with 85 % overall classification accuracy and 75 % dust detection precision. In addition, both the support vector machine and the convolutional neural network detects more dust particles (on average) than the on-board classification algorithm, with 14 % ± 1 % and 16 % ± 7 % detection enhancement respectively. The proposed convolutional neural network classifier (or similar tools) should therefore be considered for post-processing of the electric field signals observed by the Solar Orbiter.

White blood cells (WBCs) are the most diverse types of cells observed in the healing process of injured skeletal muscles. In the recovery process, WBCs exhibit a dynamic cellular response and undergo multiple changes of the protein expression. The progress of healing can be analyzed by the number of WBCs or by the number of specific proteins observed in light microscopy images obtained at different time points after injury. We propose a deep learning quantification and analysis system called DeepQuantify to analyze WBCs in light microscopy images of uninjured and injured muscles of female mice. The DeepQuantify system features in segmentation using the localized iterative Otsu's thresholding method, masking postprocessing, and classification of WBCs with a convolutional neural network (CNN) classifier to achieve a high accuracy and a low manpower cost. The proposed two-layer CNN classifier designed based on the optimization hypothesis is evaluated and compared with other CNN classifiers. The DeepQuantify system adopting these CNN classifiers is evaluated for quantifying CD68-positive macrophages and 7/4-positive neutrophils and compared with the state-of-the-art deep learning segmentation architectures. DeepQuantify achieves an accuracy of 90.64% and 89.31% for CD68-positive macrophages and 7/4-positive neutrophils, respectively. The DeepQuantify system employing the proposed two-layer CNN architecture achieves better performance than those deep segmentation architectures. The quantitative analysis of two protein dynamics during muscle recovery is also presented.

Identification of crops is an important topic in the agricultural domain. Hyperspectral remote sensing data are very useful for crop feature extraction and classification. Remote sensing data is an unstructured data and Convolutional Neural Network (CNN) can work on unstructured data efficiently. This paper presents an evaluation of CNN for crop classification using the Indian Pines standard dataset obtained from the AVIRIS sensor and the study area dataset obtained from the EO-1hyperion sensor. Optimized CNN has been tuned by training the model on different parameters. It has been compared with two classification algorithms: Deep Neural Network (DNN) and Convolutional Autoencoder. According to the test results, the proposed optimized CNN model provided better results as compared to the other two methods. CNN has given 97 ± 0.58% overall accuracy for the Indian Pines standard dataset and 78 ± 2.43% for our study area dataset.

The report presents an algorithm for constructing a convolutional hierarchical neural network classifier, which is a modification of the algorithm for constructing hierarchical neural network classifiers suggested before. The original algorithm was designed to exploit intrinsic class hierarchy to build a class tree with a neural network in each node classifying groups of initial classes (in a non-terminal node) or a subset of original classes (in a terminal node). The convolutional modification utilizes convolutional neural networks instead of regular fully connected networks in order to apply the model to image classification tasks. Use of class hierarchy for image classification should reduce the number of adjusted neural network parameters compared to deep convolutional neural networks, and therefore it should reduce training and inference time. In this context the algorithm may be compared with some pruning techniques. The convolutional hierarchical neural network classifier inherits some hyperparameters of a conventional hierarchical neural network classifier, like the activation threshold and the threshold by the share of voting patterns. The goal of this study was to explore different strategies of choosing these hyperparameters. To test these strategies, we used the CIFAR-10 dataset. Also, for demonstration purposes we apply the convolutional hierarchical neural network classifier to the CIFAR-100 dataset.

Instead of using current deep-learning segmentation models (like the UNet and variants), we approach the segmentation problem using trained Convolutional Neural Network (CNN) classifiers, which automatically extract important features from images for classification. Those extracted features can be visualized and formed into heatmaps using Gradient-weighted Class Activation Mapping (Grad-CAM). This study tested whether the heatmaps could be used to segment the classified targets. We also proposed an evaluation method for the heatmaps; that is, to re-train the CNN classifier using images filtered by heatmaps and examine its performance. We used the mean-Dice coefficient to evaluate segmentation results. Results from our experiments show that heatmaps can locate and segment partial tumor areas. But use of only the heatmaps from CNN classifiers may not be an optimal approach for segmentation. We have verified that the predictions of CNN classifiers mainly depend on tumor areas, and dark regions in Grad-CAM's heatmaps also contribute to classification.

Several Medium Access Control (MAC) and routing protocols have been developed in the last years for Underwater Wireless Sensor Networks (UWSNs). One of the main difficulties to compare and validate the performance of different proposals is the lack of a common standard to model the acoustic propagation in the underwater environment. In this paper we analyze the evolution of underwater acoustic prediction models from a simple approach to more detailed and accurate models. Then, different high layer network protocols are tested with different acoustic propagation models in order to determine the influence of environmental parameters on the obtained results. After several experiments, we can conclude that higher-level protocols are sensitive to both: (a) physical layer parameters related to the network scenario and (b) the acoustic propagation model. Conditions like ocean surface activity, scenario location, bathymetry or floor sediment composition, may change the signal propagation behavior. So, when designing network architectures for UWSNs, the role of the physical layer should be seriously taken into account in order to assert that the obtained simulation results will be close to the ones obtained in real network scenarios.

The paper investigates the impact of preliminary image restoration using a denoising autoencoder (DAE) on the classification accuracy of a convolutional neural network (CNN) under various types of noise. The relevance of the topic is due to the fact that in real conditions, optical images often contain distortions caused by changes in lighting, vibrations, camera movement, and other factors, which significantly complicates the task of object recognition. Traditional filters do not always provide sufficient cleaning quality and can lead to the loss of important structural features. In this regard, the use of deep neural networks, in particular autoencoders, is a promising direction for improving the robustness of computer vision algorithms to noise of various nature. The study uses the CIFAR-10 dataset and implements a two-component model: an autoencoder for preliminary cleaning and a CNN for classification. The trained autoencoder restores the image structure after exposure to Gaussian, impulse, Poisson, and speckle noise. Three series of experiments were conducted: classification of clean images, classification of noisy data without cleaning, and classification after preliminary restoration by the autoencoder. The results showed that CNN demonstrates an accuracy of 70.37% on clean data, but when noise is introduced, the accuracy drops to 30–59% depending on the type of distortion. After applying the autoencoder, classification accuracy increased to 56–60% for all types of noise, with the greatest improvement observed for Gaussian noise with high dispersion. The results confirm that using an autoencoder as a preliminary restoration step is an effective method for improving classification accuracy and reducing CNN vulnerability to noise. This approach provides better generalization and stability of the system, which is especially important for real-time applications—in particular, in dynamic systems, robotics, autonomous transport, and navigation systems, where the quality of optical data is often unstable. The study demonstrates the promise of integrating restoration and classification models into a single structure to improve the performance of computer vision systems in challenging conditions.

We present 97 new high-quality strong lensing candidates found in the final ∼350 deg2 that complete the full ∼1350 deg2 area of the Kilo-Degree Survey (KiDS). Together with our previous findings, the final list of high-quality candidates from KiDS sums up to 268 systems. The new sample is assembled using a new convolutional neural network (CNN) classifier applied to r-band (best-seeing) and g, r, and i color-composited images separately. This optimizes the complementarity of the morphology and color information on the identification of strong lensing candidates. We apply the new classifiers to a sample of luminous red galaxies (LRGs) and a sample of bright galaxies (BGs) and select candidates that received a high probability to be a lens from the CNN (P CNN). In particular, setting P CNN > 0.8 for the LRGs, the one-band CNN predicts 1213 candidates, while the three-band classifier yields 1299 candidates, with only ∼30% overlap. For the BGs, in order to minimize the false positives, we adopt a more conservative threshold, P CNN > 0.9, for both CNN classifiers. This results in 3740 newly selected objects. The candidates from the two samples are visually inspected by seven coauthors to finally select 97 “high-quality” lens candidates which received mean scores larger than 6 (on a scale from 0 to 10). We finally discuss the effect of the seeing on the accuracy of CNN classification and possible avenues to increase the efficiency of multiband classifiers, in preparation of next-generation surveys from ground and space.

A biosensing method for the direct serological detection of liver diseases by integrating a SERS-based sensor and a CNN classifier

Development of machine learning solutions for prediction of functional and clinical significance of cancer driver genes and mutations are paramount in modern biomedical research and have gained a significant momentum in a recent decade. In this work, we integrate different machine learning approaches, including tree based methods, random forest and gradient boosted tree (GBT) classifiers along with deep convolutional neural networks (CNN) for prediction of cancer driver mutations in the genomic datasets. The feasibility of CNN in using raw nucleotide sequences for classification of cancer driver mutations was initially explored by employing label encoding, one hot encoding, and embedding to preprocess the DNA information. These classifiers were benchmarked against their tree-based alternatives in order to evaluate the performance on a relative scale. We then integrated DNA-based scores generated by CNN with various categories of conservational, evolutionary and functional features into a generalized random forest classifier. The results of this study have demonstrated that CNN can learn high level features from genomic information that are complementary to the ensemble-based predictors often employed for classification of cancer mutations. By combining deep learning-generated score with only two main ensemble-based functional features, we can achieve a superior performance of various machine learning classifiers. Our findings have also suggested that synergy of nucleotide-based deep learning scores and integrated metrics derived from protein sequence conservation scores can allow for robust classification of cancer driver mutations with a limited number of highly informative features. Machine learning predictions are leveraged in molecular simulations, protein stability, and network-based analysis of cancer mutations in the protein kinase genes to obtain insights about molecular signatures of driver mutations and enhance the interpretability of cancer-specific classification models.

In the summer of 2021, the US Navy conducted a Full Ship Shock Trial (FSST) for the USS Gerald R. Ford. This involved three large underwater explosions off the coast of Florida, USA. We collected underwater acoustic recordings, using low-sensitivity recorders, for the Naval Undersea Warfare Center to validate their underwater acoustic propagation models. We also deployed SoundTraps on the shallow moorings to collect additional acoustical biologics data in hopes of measuring any acoustic responses of marine fauna to the explosions. The acoustic energy of the explosions did not propagate up the continental slope and were hence not captured by the SoundTraps on the shallow moorings. However, our analyses did yield some notable changes in acoustic behavior after as compared to before the explosions. Here, we describe how our field plan was designed for before-after control-impact (BACI) hypothesis testing and discuss our analyses and findings of the few significant cases. These results lend insight for improving the impact assessments and conducting behavioral response studies during a future FSST or other large underwater explosions.

The Local Climate Zone (LCZ) classification is already widely used in urban heat island and other climate studies. The current classification method does not incorporate crucial urban auxiliary GIS data on building height and imperviousness that could significantly improve urban-type LCZ classification utility as well as accuracy. This study utilized a hybrid GIS- and remote sensing imagery-based framework to systematically compare and evaluate different machine and deep learning methods. The Convolution Neural Network (CNN) classifier outperforms in terms of accuracy, but it requires multi-pixel input, which reduces the output’s spatial resolution and creates a tradeoff between accuracy and spatial resolution. The Random Forest (RF) classifier performs best among the single-pixel classifiers. This study also shows that incorporating building height dataset improves the accuracy of the high- and mid-rise classes in the RF classifiers, whereas an imperviousness dataset improves the low-rise classes. The single-pass forward permutation test reveals that both auxiliary datasets dominate the classification accuracy in the RF classifier, while near-infrared and thermal infrared are the dominating features in the CNN classifier. These findings show that the conventional LCZ classification framework used in the World Urban Database and Access Portal Tools (WUDAPT) can be improved by adopting building height and imperviousness information. This framework can be easily applied to different cities to generate LCZ maps for urban models.