Multi-scale residual attention network for seismic data interpolation

MSRConvNet: Classification of railway track defects using multi-scale residual convolutional neural network

Seismic data interpolation and denoising play vital roles in obtaining complete and clean data in seismic data processing. Seismic data usually misses along various spatial axes and always mix with random noise. In order to obtain complete and clean seismic data, reconstruction technology can interpolate missing data and attenuate random noise. Nonsubsampled contourlet transform is an effective transform to obtain multi-scale and multi-direction sparse domain data for compression sensing interpolation and denoising. However, conventional iterative shrinkage/thresholding cannot handle ill-posed and ill-conditioned equations for solving linear inverse problem. We present a two-step iterative log thresholding method to overcome ill-posed and ill-conditioned problems and improve the convergence rate and solution accuracy, which can interpolate and denoise seismic data simultaneously in the nonsubsampled contourlet transform framework. First, we use nonsubsampled contourlet transform to convert the seismic missing data with random noise to sparse domain. Then, we apply two-step iterative log thresholding algorithm to interpolate and denoise data in sparse domain. The result of each iteration is based on the results of the previous two iterations, which can accelerate convergence rate. In addition, log thresholding can further improve convergence rate and solution accuracy. Finally, we use inverse nonsubsampled contourlet transform to obtain the interpolated and denoised seismic data. The new method can reconstruct the irregularly missing data and attenuate random noise to obtain complete and clean seismic data with high accuracy, which is crucial for seismic imaging and inversion. We demonstrate the applicability and effectiveness of this simultaneous interpolation and denoising technique with successful applications to both synthetic and field data examples.

In recent years, seismic data processing based on deep convolutional neural networks (CNN) has made great progress. However, most of these methods rely on the feature information on the same scale and cannot make full use of the self-similarity of seismic data. In order to solve this problem, this paper proposes a novel Pyramid Attention Residual Neural Network (PARNet) for seismic data denoising. Specifically, the main framework of the network includes the residual block (ResBlock), the residual block with multi-core convolutional layer, the parallel space and channel attention (MSCARB) and the pyramid module(Pyramid Module). Among them, MSACRB can not only extract more abundant features, but also focus on the features of channel and spatial dimension, so as to achieve stronger feature representation. The pyramid module captures multi-scale features through dilated convolution with different expansion rates. At the same time, the global context module can capture the global information of the feature map. The combination of the above two modules can achieve the purpose of capturing multi-scale global context features. This method has been verified on synthetic seismic data and field seismic data. The experiments use PSNR and SSIM as evaluation indicators. A large number of experiments have demonstrated that PARNet has efficient denoising ability and a competitive advantage compared with the latest seismic data denoising methods.

Seismic data interpolation is a longstanding issue. Most current methods are only suitable for randomly missing cases. To deal with regularly missing cases, an antialiasing strategy should be included. However, seismic survey design using a random distribution of shots and receivers is always operationally challenging and impractical. We have used deep-learning-based approaches for seismic data antialiasing interpolation, which could extract deeper features of the training data in a nonlinear way by self-learning. It can also avoid linear events, sparsity, and low-rank assumptions of the traditional interpolation methods. Based on convolutional neural networks, eight-layers residual learning networks (ResNets) with a better back-propagation property for deep layers is designed for interpolation. Detailed training analysis is also performed. A set of simulated data is used to train the designed ResNets. The performance is assessed with several synthetic and field data. Numerical examples indicate that the trained ResNets can help to reconstruct regularly missing traces with high accuracy. The interpolated results in the time-space domain and the frequency-wavenumber ([Formula: see text]-[Formula: see text]) domain demonstrate the validity of the trained ResNets. Even though the accuracy decreases with the increase of the feature difference between the test and training data, the proposed method can still provide reasonable interpolation results. Finally, the trained ResNets is used to reconstruct dense data with halved trace intervals for synthetic and field data. The reconstructed dense data are more continuous along the spatial direction, and the spatial aliasing effects disappear in the [Formula: see text]-[Formula: see text] domain. The reconstructed dense data have the potential to improve the accuracy of subsequent seismic data processing and inversion.

Seismic data interpolation is essential in a seismic data processing workflow, recovering data from sparse sampling. Traditional and deep-learning-based methods have been widely used in the seismic data interpolation field and have achieved remarkable results. In this paper, we develop a seismic data interpolation method through the novel application of diffusion probabilistic models (DPMs). DPM transforms the complex end-to-end mapping problem into a progressive denoising problem, enhancing the ability to reconstruct complex situations of missing data, such as large proportions and large-gap missing data. The interpolation process begins with a standard Gaussian distribution and seismic data with missing traces and then removes noise iteratively with a UNet trained for different noise levels. Our DPM-based interpolation method allows interpolation for various missing cases, including regularly missing, irregularly missing, consecutively missing, noisy missing, and different ratios of missing cases. The ability to generalize to different seismic data sets is also discussed in this paper. Numerical results of synthetic and field data indicate satisfactory interpolation performance of the DPM-based interpolation method in comparison with the f-x prediction filtering method, the curvelet transform method, the low-dimensional manifold method (LDMM), and the coordinate attention-based UNet method, particularly in cases with large proportions and large-gap missing data. Diffusion is all we need for seismic data interpolation.

Computer-aided diagnosis is of great significance to improve the diagnostic accuracy of pancreatic cancer that has an insidious course and does not have obvious symptoms at first. However, segmentation of pancreatic cancer is challenging because the tumors vary in size with the smallest tumor having a size of about 0.5 in diameter, and most of them have irregular shapes and unclear boundaries. In this study, we developed a deep learning architecture Multi-Scale Channel Attention Unet (MSCA-Unet) for pancreatic tumor segmentation and collected CT images of 419 patients from The Affiliated Hospital of Qingdao University and a public dataset. We embedded the multi-scale network into the encoder to extract semantic information at different scales and the decoder to provide supplemental information to overcome the loss of information in the upsampling and the drift of the localized tumor due to the upsampling and skip connections. We adopted the channel attention unit after the multi-scale convolution to emphasize the informative channels, which was observed to have the effects of accelerating the positioning process, reducing false positives, and improving the accuracy of outlining very small, irregular pancreatic tumors. Our results show that our network outperformed the other current mainstream segmentation networks and achieved a Dice index of 68.03%, a Jaccard of 59.31%, and an FPR of 1.36% on the private dataset Task-01 without data pre-processing. Compared with the other pancreatic tumor segmentation networks on the public dataset Task-02, our network produced the best Dice index, 80.12%, with the assistance of the data pre-processing scheme. This study strategically utilizes the multi-scale convolution and channel attention mechanism of the architecture to provide a dedicated network for segmentation of small and irregular pancreatictumors.

Diabetic retinopathy grading based on multi-scale residual network and cross-attention module

This paper addresses the central issue arising from the (SDAC) of facial expressions, namely, to balance the competing demands of good global features for detection, and fine features for good facial expression classifications by replacing the feature extraction part of the "neck" network in the feature pyramid network in the You Only Look Once X (YOLOX) framework with a novel architecture involving three attention mechanisms-batch, channel, and neighborhood-which respectively explores the three input dimensions-batch, channel, and spatial. Correlations across a batch of images in the individual path of the dual incoming paths are first extracted by a self attention mechanism in the batch dimension; these two paths are fused together to consolidate their information and then split again into two separate paths; the information along the channel dimension is extracted using a generalized form of channel attention, an adaptive graph channel attention, which provides each element of the incoming signal with a weight that is adapted to the incoming signal. The combination of these two paths, together with two skip connections from the input to the batch attention to the output of the adaptive channel attention, then passes into a residual network, with neighborhood attention to extract fine features in the spatial dimension. This novel dual path architecture has been shown experimentally to achieve a better balance between the competing demands in an SDAC problem than other competing approaches. Ablation studies enable the determination of the relative importance of these three attention mechanisms. Competitive results are obtained on two non-aligned face expression recognition datasets, RAF-DB and SFEW, when compared with other state-of-the-art methods.

Energy disaggregation technology is a key technology to realise real-time non-intrusive load monitoring (NILM). Current energy disaggregation methods use the same scale to extract features from the sequence, which makes part of the local features lost, resulting in low recognition accuracy of electrical appliances. Aiming at the low accuracy of non-intrusive energy disaggregation with low-frequency sampling, a sequential multi-scale residual network is proposed. Multi-scale residual network extracts multi-scale feature information through multi-scale convolution, and uses residual learning to deepen the network structure to further improve the performance of the algorithm. Different scales features are captured by multi-scale convolution to improve the recognition accuracy of electrical appliances with low using frequency. Sequence-to-sequence energy disaggregation mechanism can improve the disaggregation efficiency of the algorithm. The experimental comparison results show that the model can get better disaggregation effect, and can effectively identify the start-stop state of electrical appliances.

ResNet and its application to medical image processing: Research progress and challenges

ABSTRACTInterpolation techniques provide an effective method for recovery of missing traces. In recent years, many researchers have applied deep learning methods to seismic data interpolation. Generally, one can choose synthetic data as a training set; however, the features of synthetic data are always inconsistent with those of field data, which may lead to inaccurate interpolation. Meanwhile, U‐Net is a common network structure used in seismic data interpolation; however, the four downsampling and upsampling structures of U‐Net have limited adaptability for different data. In this study, the deep learning method based on U‐Net++ was proposed for seismic data interpolation, which contains U‐Net with different depths. The different depths were connected by skip pathways, and the best depth of the network was chosen for different seismic data by deep supervision. Furthermore, a new strategy for training sets was designed: frequency‐wavenumber (f‐k) bandpass filters were used to convert natural images into a natural seismic training set, which has a stronger generalization capability than synthetic data as the training set. The characteristics of the new training set can effectively improve the accuracy of missing data reconstruction. Compared with the conventional U‐Net and traditional interpolation techniques, for example, the Fourier Bregman method, the proposed method produces more accurate and reasonable interpolation results. Further, it can reconstruct both irregular and regular missing seismic data, even in the presence of strong random noise and aliasing. Synthetic and field data tests showed the effectiveness, robustness and generalization of the proposed method.

Fault diagnosis of drainage pumps is essential for ensuring the safe and stable operation of drainage systems. However, vibration signals acquired from drainage pumps under practical operating conditions are often characterized by strong non-stationarity, unavoidable environmental interference, and fault features distributed across multiple temporal scales, which poses significant challenges to accurate multi-class fault identification. To tackle these challenges, this paper proposes a multi-scale residual attention convolutional neural network (MRANet) for drainage pump fault diagnosis. A multi-scale residual attention feature extraction (MRAFE) module is designed to jointly model fault-related information at different temporal scales by means of parallel convolutions and residual connections, while attention mechanisms are incorporated within each scale branch to enhance discriminative fault features. Furthermore, a complementary multi-level feature integration (CMFI) strategy is developed to effectively integrate features extracted at different network depths, enabling end-to-end fault identification directly from raw one-dimensional vibration signals. Extensive experiments are conducted on a drainage pump vibration dataset containing eight operating conditions. Experimental results show that MRANet achieves an accuracy of 95.43%, outperforming several benchmark models. These results verify the effectiveness and robustness of the proposed method for multi-class fault diagnosis of drainage pumps.

Summary The need for accurate and efficient seismic data interpolation is increasing. Insufficient spatial sampling may bring negative effects on existing migration and inversion methods. In recent years, Deep Learning (DL) techniques have shown great potential for various tasks. It can extract high-level features from data through self-learning mechanism. In this paper, we focus on severe under-sampled seismic data interpolation problem (scale of 4) and develop two Convolutional Neural Networks (CNN) based models. We build a deep cascaded interpolation network (DCIN) as a baseline model. To get better performance and avoid per-interpolation, we develop a modified deep residual interpolation network (MDRIN). Examples of two synthetic datasets demonstrate that our networks can obtain much higher performance and possess anti-aliasing features. For new dataset, we introduce the strategy of transfer learning, which can get comparable performance via using a very small amount of data compared to training the network from scratch by using the full amount of data. Integration of the MDRIN model and transfer learning strategy shows high potential for application in field data.

Image dehazing is a useful technique which can eliminate the bad effect of haze on images and enhance the performances of image/video processing algorithms in hazy weather. In this chapter, we first propose a single-image dehazing method. We estimate the initial transmission properly based on latent region segmentation and refine the estimated initial transmission by an objective function with a novel weighted L1-norm regularization term. The half-quadratic splitting minimization method is employed to solve this optimization problem. We also define an evaluation function to estimate the reliable global atmospheric light. With the refined transmission map and atmospheric light, we recover the haze-free image by the haze imaging model. We then propose a retinex-based decomposition model for a hazy image and a novel end-to-end image dehazing network. In the model, the illumination of the hazy image is decomposed into natural illumination for the haze-free image and residual illumination caused by haze. Based on this model, we design a deep retinex dehazing network (RDN) to jointly estimate the residual illumination map and the haze-free image. Our RDN consists of a multiscale residual dense network and a U-Net with channel and spatial attention mechanisms. The multiscale residual dense network can simultaneously capture global contextual information from small-scale receptive fields and local detailed information from large-scale receptive fields to precisely estimate the residual illumination map caused by haze. And in the dehazing U-Net, we apply the channel and spatial attention mechanisms in the skip connection to achieve a trade-off between overdehazing and underdehazing by automatically adjusting the channel-wise and pixel-wise attention weights so as to obtain a refined haze-free image.

Accurate fault diagnosis is the key to ensure the safe and operation of rotating machines. Accuracy of bearing fault diagnosis will be reduced in complex noise environments and variable load conditions. Therefore, this paper proposes a fault diagnosis method based on attention mechanism and residual neural network (FaultCNN). Firstly, a residual attention module is designed to extract the original fault features of the noise signals. Secondly, the ability to extract features of the residual module is further enhanced by adding an average pooling layer to the residual structure shortcut. To validate the diagnostic performance of the model, CWRU datasets are used for experimental verification. In the noise environment of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\text{SNR}=-6 \text{dB}$</tex> , the accuracy of model fault diagnosis is 94.23%. Compared to existing bearing fault diagnostic methods, FaultCNN has higher anti-noise performance and variable load adaptability.