Search

Extracting balanced geostrophic motions (BM) from sea surface height (SSH) observations obtained by wide‐swath altimetry holds great significance in enhancing our understanding of oceanic dynamic processes at submesoscale wavelength. However, SSH observations derived from wide‐swath altimetry are characterized by high spatial resolution while relatively low temporal resolution, thereby posing challenges to extract the BM from a single SSH snapshot. To address this issue, this paper proposes a deep learning model called the BM‐UBM Network, which takes an instantaneous SSH snapshot as input and outputs the projection corresponding to the BM. Training experiments are conducted both in the Gulf Stream and South China Sea, and three metrics are considered to diagnose model's outputs. The favorable results highlight the potential capability of the BM‐UBM Network to process SSH measurements obtained by wide‐swath altimetry. Plain Language Summary: Oceanic dynamic processes can be classified into two categories: balanced geostrophic motions (BM), including large‐scale circulation, mesoscale and submesoscale eddy turbulence, and unbalanced wave motions (UBM), including barotropic tides, and inertia–gravity waves (IGWs). Both types of motions coexist and have respective contributions to the sea surface height (SSH). How to extract the BM from the total SSH observations obtained by satellite altimetry is the crucial problem to be solved in this paper. To tackle this issue, we propose a deep learning model named the BM‐UBM Network to establish the relationship between the total SSH and the BM component. The BM‐UBM Network can generate SSH estimations for the BM when provided with a well‐resolved SSH snapshot. Key Points: A Deep learning model is developed to extract balanced motions from sea surface height snapshot based on a realistic simulationDiagnostics of three metrics reveal the effectiveness of the model in extracting balanced motionsThe model exhibits remarkable advantages over the Gaussian filter (baseline) in capturing the gradient and Laplacian information [ABSTRACT FROM AUTHOR]

Summary: Conventional fault recognition algorithms can only recognize the classes of faults that have appeared in wind turbine systems. However, if a new category of faults appears, the traditional algorithm can only misclassify it into the class of pre‐existing faults. In this paper, a new class fault recognition method based on deep learning is proposed. Firstly, the initialized model is built using known fault data, including detectors and classifiers. Secondly, the new class fault data detected by the detectors are put into the cache, and when the cache overflows, the new class faults are augmented with data using generative adversarial networks. Finally, the new class fault data and the generated data are added to the initial training set, and the structure and weights of the initialized model are updated using the new training set. In this way, the purpose of recognizing the new class of faults is achieved. The experimental results show that the proposed detection algorithm can effectively detect new class faults and the model can efficiently solve the problem of the new class of fault recognition after data augmentation. [ABSTRACT FROM AUTHOR]

The evolution of deep learning (DL) has made artificial intelligence image recognition a mature technology. Recently, the use of DL to identify diabetic retinopathy (DR) has been recognized as a major challenge. Retinal abnormalities caused by DR can damage the retina and thus cause permanent damage or even blindness. Therefore, the detection of diabetes symptoms at an early stage can help to considerably reduce the risk of blindness. The differences in environments, equipment, and photographers have led to the inconsistent specifications of images and have thus affected the efficiency of the training model for classifying the DR level. If low‐quality fundus images are removed through an image quality assessment (Eye‐Quality Library, EyeQ), the number of images of the trained model would considerably reduce, in turn affecting the reliability of the training model. To solve this problem, this paper proposes a preprocessing method to strengthen the image features. The results obtained in this study revealed that the preprocessing method could increase the amount of data available for the training model. Thus, this study improved the EfficientNet model for the enhancement of the classification performance of the DR level. The results also showed that an increase in the model accuracy from 0.7727 to 0.7920 for the classification of the different stages of DR. In addition, the results revealed that the revised EfficientNet could obtain better average area under the ROC curve among the five classes (0.926) than MobileNet (0.54) and the original EfficientNet (0.922). Finally, this study implemented the proposed system by using an application programming interface (API) to enable the users to upload a fundus image to the API and obtain the DR results. [ABSTRACT FROM AUTHOR]

SUMMARY: Plant diseases worsen the threat of food shortage with the growing global population, and disease recognition is the basis for the effective prevention and control of plant diseases. Deep learning has made significant breakthroughs in the field of plant disease recognition. Compared with traditional deep learning, meta‐learning can still maintain more than 90% accuracy in disease recognition with small samples. However, there is no comprehensive review on the application of meta‐learning in plant disease recognition. Here, we mainly summarize the functions, advantages, and limitations of meta‐learning research methods and their applications for plant disease recognition with a few data scenarios. Finally, we outline several research avenues for utilizing current and future meta‐learning in plant science. This review may help plant science researchers obtain faster, more accurate, and more credible solutions through deep learning with fewer labeled samples. Significance Statement: This paper focuses on the research related to meta‐learning based plant disease recognition with few samples. With the significant progress in deep‐learning based plant disease detection, disease diagnose is gradually intelligent. However, traditional deep‐learning has limited the existing research development because of the great demand for samples, time and effort consumption, and the lack of migration between models. Furthermore, traditional deep‐learning faces the time‐consuming and labor‐intensive nature of the available samples and the lack of transferability between models. The emergence of meta‐learning methods has solved the problem of sample annotation arising from the huge sample size, and strong feature‐extracting ability, which may be an essential tool for future crop phenotype research. [ABSTRACT FROM AUTHOR]

Summary: In the presence of noise in communication systems, constellation diagram points are scattered to the extent that may make the modulation classification a difficult task. With the plethora of applications of machine and deep learning, several communication systems have adopted machine and deep learning to solve some classical detection and classification problems. Casting the modulation order detection as a pattern classification of the constellation images opens the door for application of mature machine learning and image processing tools to solve the classification problem, efficiently. This paper presents a system based on a wavelet‐aided convolutional neural network (CNN) classifier to efficiently detect the modulation type and order in the presence of noise. The proposed system depends on a pretrained CNN setup, which is trained with a set of constellation diagrams for each modulation scheme and used after that for testing. In addition, discrete wavelet transform (DWT) is investigated to generate representative patterns from constellation diagrams to be used for the training and testing tasks as well. The wavelet approximation images and their corresponding wavelet sub‐bands across all predefined scales are used in the dataset. Several pretrained networks including AlexNet, VGG‐16, and VGG‐19 are used as classifiers for the modulation type from the DWTs for different constellation diagrams. Several simulation experiments are presented in this paper to compare different scenarios for modulation classification at different signal‐to‐noise ratios (SNRs). [ABSTRACT FROM AUTHOR]

A coreset of an input set is its small summarization, such that solving a problem on the coreset as its input, provably yields the same result as solving the same problem on the original (full) set, for a given family of problems (models/classifiers/loss functions). Coresets have been suggested for many fundamental problems, for example, in machine/deep learning, computer vision, databases, and theoretical computer science. This introductory paper was written following requests regarding the many inconsistent coreset definitions, lack of source code, the required deep theoretical background from different fields, and the dense papers that make it hard for beginners to apply and develop coresets. The article provides folklore, classic, and simple results including step‐by‐step proofs and figures, for the simplest (accurate) coresets. Nevertheless, we did not find most of their constructions in the literature. Moreover, we expect that putting them together in a retrospective context would help the reader to grasp current results that usually generalize these fundamental observations. Experts might appreciate the unified notation and comparison table for existing results. Open source code is provided for all presented algorithms, to demonstrate their usage, and to support the readers who are more familiar with programming than mathematics. This article is categorized under:Algorithmic Development > Structure DiscoveryFundamental Concepts of Data and Knowledge > Big Data MiningTechnologies > Machine Learning [ABSTRACT FROM AUTHOR]

A polyp is one of the major causes of gastroenterology, which leads to colorectal cancer. The detection of polyps by colonoscopy imaging is a significant challenge because of the diversity in polyp structure and lack of examination accuracy. To solve this problem, the automatic segmentation of polyps can be used to enhance examination accuracy and reduce gastrointestinal (GI) disease. In this paper, the framework of polyp image segmentation is developed by a deep learning approach, especially a convolutional neural network. This proposed framework used the Kvasir‐SEG database, which contains 1000 GI polyp images and corresponding segmentation masks according to annotation by medical experts. This database is divided into 900 for training images and 100 for testing images. This framework is based on image preprocessing and two types of SegNet architecture to obtain the segmented polyp image. This paper has demonstrated state‐of‐the‐art performance on both VGG‐16, and VGG‐19 networks for training and testing data to address colorectal cancer screening rates. The results confirmed that the VGG‐19 model has outperformed the VGG‐16 model via all evaluation parameters except sensitivity for the polyp segmentation on the Kvasir‐SEG dataset. Additionally, it will support gastroenterologists during medical strategy to correctly choose the treatment with less time. [ABSTRACT FROM AUTHOR]

In this paper, a novel deep learning (DL) approach is developed to solve the electromagnetic inverse scattering (EMIS) problems. Many challenges, such as ill‐posedness, high computational cost, and strong non‐linearity, are encountered when solving the EMIS problems. To surmount these difficulties, a multi‐model fusion convolutional neural network architecture is proposed, termed here as Amplitude‐Phase scheme. To the best of our knowledge, it is the first time that the multi‐model fusion DL approach is employed to solve the EMIS problems. Amplitude data and phase data of the measured scattering data are applied to train the proposed scheme. Furthermore, we compare APs with three different training schemes, including Amplitude‐Only scheme, and Phase‐Only scheme, and Complex‐Value scheme. The performance of the proposed DL schemes has been validated by numerical simulations. The results demonstrate that the proposed multi‐model fusion approach outperforms other DL schemes in terms of accuracy and is able to achieve a better performance in reconstructing homogeneous and heterogeneous scatterers. [ABSTRACT FROM AUTHOR]

The development of deep learning provides a new way for solving the colorization problem on the grayscale image. Excellent coding‐based methods appear in the automatic image colorization task, avoiding the unsaturated colour effect problem of previous methods based on the L2 loss function. Traditional neural networks come with high computational costs and a large number of parameters. Considering the limitation of memory and computing resources and aiming at lightweight, a novel grey image automatic colorization network is proposed. The basic idea of coding‐based methods is used, regarding the colorization task as a pixel‐level classification problem, meanwhile redesign and improve the colour encoding and decoding process. This network architecture leverages a lightweight convolution to reduce the computation and combines an efficient attention model to form a residual block as the kernel of the backbone network. Furthermore, an efficient image self‐attention mechanism placed at the end of the network is applied to enhance the ultimate colouring results. The method proposed in this paper can maintain the natural colouring effect and significantly reduce the computational amount and network model parameters. [ABSTRACT FROM AUTHOR]

Visual inspection of photovoltaic modules using electroluminescence (EL) images is a common method of quality inspection. Because human inspection requires a lot of time, object detection algorithm to replace human inspection is a popular research direction in recent years. To solve the problem of low accuracy and slow speed in EL image detection, we propose a YOLO‐based object detection algorithm YOLO‐PV, which achieves 94.55% of AP (average precision) on the photovoltaic module EL image data set, and the interference speed exceeds 35 fps. The improvement of speed and accuracy benefits from the targeted design of the network architecture according to the characteristics of EL image. First, we weaken the backbone's ability to extract deep‐level information so that it can focus on extracting the low‐level defect information. Second, the PAN network is used for feature fusion in the Neck part. But, only the single‐size feature map output is retained, which significantly reduces the amount of calculation. Also, we analyze the impact of data enhancement methods on model overfitting and performance. Finally, we give effective data enhancement methods. The results show that the object detection algorithm in this paper can meet the requirements for high‐precision and real‐time processing on the PV module production line. [ABSTRACT FROM AUTHOR]

Purpose: Radiotherapy is one of the main treatments of nasopharyngeal cancer (NPC) and lung cancer. Accurate segmentation of organs at risks (OARs) in CT images is a key step in radiotherapy planning for NPC and lung cancer. However, the segmentation of OARs is influenced by the highly imbalanced size of organs, which often results in very poor segmentation results for small and difficult‐to‐segment organs. In addition, the complex morphological changes and fuzzy boundaries of OARs also pose great challenges to the segmentation task. In this paper, we propose a cross‐layer attention fusion network (CLAF‐CNN) to solve the problem of accurately segmenting OARs. Methods: In CLAF‐CNN, we integrate the spatial attention maps of the adjacent spatial attention modules to make the segmentation targets more accurately focused, so that the network can capture more target‐related features. In this way, the spatial attention modules in the network can be learned and optimized together. In addition, we introduce a new Top‐K exponential logarithmic Dice loss (TELD‐Loss) to solve the imbalance problem in OAR segmentation. The TELD‐Loss further introduces a Top‐K optimization mechanism based on Dice loss and exponential logarithmic loss, which makes the network pay more attention to small organs and difficult‐to‐segment organs, so as to enhance the overall performance of the segmentation model. Results: We validated our framework on the OAR segmentation datasets of the head and neck and lung CT images in the StructSeg 2019 challenge. Experiments show that the CLAF‐CNN outperforms the state‐of‐the‐art attention‐based segmentation methods in the OAR segmentation task with average Dice coefficient of 79.65% for head and neck OARs and 88.39% for lung OARs. Conclusions: This work provides a new network named CLAF‐CNN which contains cross‐layer spatial attention map fusion architecture and TELD‐Loss for OAR segmentation. Results demonstrated that the proposed method could obtain accurate segmentation results for OARs, which has a potential of improving the efficiency of radiotherapy planning for nasopharynx cancer and lung cancer. [ABSTRACT FROM AUTHOR]

Purpose: The research is to improve the efficiency and accuracy of recognition of honeycomb lung in CT images. Methods: Deep learning methods are used to achieve automatic recognition of honeycomb lung in CT images, however, are time consuming and less accurate due to the large amount of structural parameters. In this paper, a novel recognition method based on MobileNetV1 network, multiscale feature fusion method (MSFF), and dilated convolution is explored to deal with honeycomb lung in CT image classification. Firstly, the dilated convolution with different dilated rate is used to extract features to obtain receptive fields of different sizes, and then fuse the features of different scales at multiscale feature fusion block is used to solve the problem of feature loss and incomplete feature extraction. After that, by using linear activation functions (Sigmoid) instead of nonlinear activation functions (ReLu) in the improved deep separable convolution blocks to retain the feature information of each channel. Finally, by reducing the number of improved deep separable blocks to reduce the computation and resource consumption of the model. Results: The experimental results show that improved MobileNet model has the best performance and the potential for recognition of honeycomb lung image datasets, which includes 6318 images. By comparing with 4 traditional models (SVM, RF, decision tree, and KNN) and 11 deep learning models (LeNet‐5, AlexNet, VGG‐16, GoogleNet, ResNet18, DenseNet121, SENet18, InceptionV3, InceptionV4, Xception, and MobileNetV1), our model achieved the performance with an accuracy of 99.52%, a sensitivity of 99.35%, and a specificity of 99.89%. Conclusion: Improved MobileNet model is designed for the automatic recognition and classification of honeycomb lung in CT images. Through experiments comparative analysis of other models of machine learning and deep learning, it is proved that the proposed improved MobileNet method has the best recognition accuracy with fewer the model parameters and less the calculation time. [ABSTRACT FROM AUTHOR]