Facies-Guided Seismic Image Super-Resolution

Seismic data is an essential source of information often contaminated with disturbing, coherent and random noise. Seismic random noise has degenerative impacts on subsequent seismic processing and data interpretation. Thus, seismic noise attenuation is a key step in seismic processing. Convolutional Neural Networks (CNNs) have proven successful for various image processing tasks in multidisciplinary fields and this paper aims to study the impact of three CNN architectures (autoencoders, denoising CNNs (DnCNN) and residual dense networks (RDN)) on improving the signal to noise ratio of seismic data. The work consists of three steps: Data preparation, model training and model testing. In this study we have used real seismic data to prepare the training dataset we have manually added noise. Most studies on seismic noise attenuation, study only a single kind of noise. However this paper suggests making our approach by exposing the model to many kinds of noises and noise levels such as Guassian noise, Poisson noise, Salt and Pepper and Speckle noise. In this paper we have analysed the performance of three models. Autoencoders: This architecture consists of two parts, the encoders and the decoders. The encoder consists of convolutions on the input image to extract all key information and map it to a latent space with loss of unnecessary data(noise) while the decoder reconstructs the image from the latent space to a seismic image while high signal to noise ratio. DnCNNs: This architecture is a combination of residual learning and batch normalization and mainly consists of three kinds of blocks. The model is trained to predict the residual image, that is the difference between the noisy observation and the latent clean image. RDNs: This architecture comprises of shallow feature extraction net, residual dense blocks (RDBs), dense feature fusion, and lastly up-sampling net. The data prepared as mentioned above is trained on all three CNN models across different noise levels and the performance of these models was compared. The model is finally tested on a batch of unseen noisy seismic sections and the performance is measured by an l2 loss namely mean squared error and the improvement in signal to noise ratio. The resultant images from all three architectures across different noise levels have drastically improved signal to noise ratio and thus the application of CNN as a denoiser for seismic images proves to be successful. It is important to note that when comparing the difference plots(Noisy image minus the denoised image) we found minimal signal leakage. While the application of CNN for image pre-processing has seen great success in other fields, mathematical denoising techniques such as F-K filter, tao-p filter are still used in oil and gas industry particularly in seismic denoising. After thorough review, this paper studies some of the most successful denoising CNN architectures and its success in seismic denoising.

In this paper, we propose to address the downscaling of ocean remote sensing data using image super-resolution models based on deep learning, and more particularly Convolutional Neural Networks (CNNs). The goal of this study, for which we focus on satellite-derived Sea Surface Temperature (SST) data, is to evaluate the efficiency and the relevance of deep learning architectures applied to oceanographic remote sensing data. By using a CNN architecture, namely SRCNN (Super Resolution CNN), on a large-scale dataset of SST fields, we show that it allows a considerable gain in terms of PSNR compared to classical downscaling techniques. These results point out the relevance of deep learning models specifically trained for ocean remote sensing data and advocate for other applications to the reconstruction of high-resolution sea surface geophysical fields from multi-sensor satellite observations.

Convolutional neural networks (CNN) applications range from recognition and reasoning (such as handwriting recognition, facial expression recognition and video surveillance) to intelligent text applications such as semantic text analysis and natural language processing applications. Two key observations drive the design of a new architecture for CNN. First, CNN workloads exhibit a widely varying mix of three types of parallelism: parallelism within a convolution operation, intra-output parallelism where multiple input sources (features) are combined to create a single output, and inter-output parallelism where multiple, independent outputs (features) are computed simultaneously. Workloads differ significantly across different CNN applications, and across different layers of a CNN. Second, the number of processing elements in an architecture continues to scale (as per Moore's law) much faster than off-chip memory bandwidth (or pin-count) of chips. Based on these two observations, we show that for a given number of processing elements and off-chip memory bandwidth, a new CNN hardware architecture that dynamically configures the hardware on-the-fly to match the specific mix of parallelism in a given workload gives the best throughput performance. Our CNN compiler automatically translates high abstraction network specification into a parallel microprogram (a sequence of low-level VLIW instructions) that is mapped, scheduled and executed by the coprocessor. Compared to a 2.3 GHz quad-core, dual socket Intel Xeon, 1.35 GHz C870 GPU, and a 200 MHz FPGA implementation, our 120 MHz dynamically configurable architecture is 4x to 8x faster. This is the first CNN architecture to achieve real-time video stream processing (25 to 30 frames per second) on a wide range of object detection and recognition tasks.

Classification task is one of the most common tasks in machine learning. This supervised learning problem consists in assigning each input to one of a finite number of discrete categories. Classification task appears naturally in numerous applications, such as medical image processing, speech recognition, maintenance systems, accident detection, autonomous driving etc.In the last decade methods of deep learning have proven to be extremely efficient in multiple machine learning problems, including classification. Whereas the neural network architecture might depend a lot on data type and restrictions posed by the nature of the problem (for example, real-time applications), the process of its training (i.e. finding model’s parameters) is almost always presented as loss function optimization problem.Cross-entropy is a loss function often used for multiclass classification problems, as it allows to achieve high accuracy results.Here we propose to use a generalized version of this loss based on Renyi divergence and entropy. We remark that in case of binary labels proposed generalization is reduced to cross-entropy, thus we work in the context of soft labels. Specifically, we consider a problem of image classification being solved by application of convolution neural networks with mixup regularizer. The latter expands the training set by taking convex combination of pairs of data samples and corresponding labels. Consequently, labels are no longer binary (corresponding to single class), but have a form of vector of probabilities. In such settings cross-entropy and proposed generalization with Renyi divergence and entropy are distinct, and their comparison makes sense.To measure effectiveness of the proposed loss function we consider image classification problem on benchmark CIFAR-10 dataset. This dataset consists of 60000 images belonging to 10 classes, where images are color and have the size of 32×32. Training set consists of 50000 images, and the test set contains 10000 images.For the convolution neural network, we follow [1] where the same classification task was studied with respect to different loss functions and consider the same neural network architecture in order to obtain comparable results.Experiments demonstrate superiority of the proposed method over cross-entropy for loss function parameter value α < 1. For parameter value α > 1 proposed method shows worse results than cross-entropy loss function. Finally, parameter value α = 1 corresponds to cross-entropy.

As crucial storage spaces for underground hydrocarbon resources, karst caves are essential for the exploitation and development of carbonate-prone reservoirs. The precise characterization of karst caves enables efficient resource management and enhances the effectiveness of oil and gas exploration and development. Due to noise interference and discontinuities in geologic strata, traditional methods that rely on seismic properties to detect karst caves are highly subjective, prone to multiple solutions, and require extensive manual intervention, resulting in low development efficiency. With the growing demand for exploration, overcoming the limitations of traditional methods through more efficient and precise karst cave identification technologies has emerged as a critical focus in carbonate-prone reservoir research. To address these limitations in 3D karst cave identification, we introduce a variant architecture of TransUnet, which integrates transformer and dual attention mechanisms with a convolutional neural network while upgrading the convolutional module from two dimensions to three dimensions to describe better the location and boundary characteristics of the karst caves in field seismic data. Compared with traditional U-shaped architectures, the model offers a faster convergence rate (epochs <10) during training and fewer false predictions during validation with open-source synthetic data. Further applications of physical models and field seismic data provide more accurate details on the positions and boundaries of karst caves, demonstrating their significant potential to improve the accuracy and resolution of karst cave detection and effectively enhance the characterization of carbonate reservoirs.

Deep learning has been successfully applied to image denoising. In this study, we take one step forward by using deep learning to suppress random noise in poststack seismic data from the aspects of network architecture and training samples. On the one hand, poststack seismic data denoising mainly aims at 3-D seismic data. We designed an end-to-end 3-D denoising convolutional neural network (3-D-DnCNN) that takes raw 3-D cubes as input in order to better extract the features of the 3-D spatial structure of poststack seismic data. On the other hand, denoising images with deep learning require noisy-clean sample pairs for training. In the field of seismic data processing, researchers usually try their best to suppress noise by using complex processes that combine different methods, but clean labels of seismic data are not available. In addition, building training samples in field seismic data has become an interesting but challenging problem. Therefore, we propose a training sample selection method that contains a complex workflow to produce comparatively ideal training samples. Experiments in this study demonstrate that deep learning can directly learn the ability to denoise field seismic data from selected samples. Although the building of the training samples may occur through a complex process, the experimental results of synthetic seismic data and field seismic data show that the 3-D-DnCNN has learned the ability to suppress the Gaussian noise and super-Gaussian noise from different training samples. Moreover, the 3-D-DnCNN network has better denoising performance toward arc-like imaging noise. In addition, we adopt residual learning and batch normalization in order to accelerate the training speed. After network training is satisfactorily completed, its processing efficiency can be significantly higher than that of conventional denoising methods.

Learning based single image super-resolution (SISR) for real-world images has been an active research topic yet a challenging task, due to the lack of paired low-resolution (LR) and high-resolution (HR) training images. Most of the existing unsupervised real-world SISR methods adopt a two-stage training strategy by synthesizing realistic LR images from their HR counterparts first, then training the super-resolution (SR) models in a supervised manner. However, the training of image degradation and SR models in this strategy are separate, ignoring the inherent mutual dependency between downscaling and its inverse upscaling process. Additionally, the ill-posed nature of image degradation is not fully considered. In this paper, we propose an image downscaling and SR model dubbed as SDFlow, which simultaneously learns a bidirectional many-to-many mapping between real-world LR and HR images unsupervisedly. The main idea of SDFlow is to decouple image content and degradation information in the latent space, where content information distribution of LR and HR images is matched in a common latent space. Degradation information of the LR images and the high-frequency information of the HR images are fitted to an easy-to-sample conditional distribution. Experimental results on real-world image SR datasets indicate that SDFlow can generate diverse realistic LR and SR images both quantitatively and qualitatively.

Fault detection is an essential component of seismic interpretation and plays a crucial role in industrial processes. However, it is also one of the main challenges, especially in delineating faults in 3D seismic data. Recently, the rapidly developing technology, deep learning, has proven to be a powerful tool for this task. A number of neural networks have been proposed for this purpose by regarding 3D fault detection as a semantic segmentation task. To further enhance the effectiveness of the deep learning methods, we propose a novel network architecture, named R2SE-Unet, to solve the 3D segmentation problem. In the neural network, we design a recurrent residual-SE convolution unit (RRCU-SE) that integrates the residual learning and Squeeze-Excitation module to store information in 3D seismic data. This component promotes the spread of 3D volumetric information and aids in learning spatial dependencies in 3D images. In addition, to reduce the impact of insufficient spatial resolution resulting from the base architecture of U-net, we add an attention unit between skip connection operations. These two new units enable our R2SE-Unet to exploit semantic information more accurately in the feature maps. After many experiments on region-based loss functions and distribution-based loss functions, we also propose a novel loss function, which takes the advantage of generalized dice (GDice) loss and balanced binary cross entropy (b-BCE) loss, named Gdice-bce, to effectively train R2SE-Unet. Although only synthetic seismic data samples are used to train the network parameters, our R2SE-Unet could produce more reliable fault feature maps on field seismic data than two other conventional fault detection neural networks. Thus, the proposed neural network is easy to train and reliably works for seismic fault interpretation on field seismic data.

The Hausdorff Distance (HD) is widely used in evaluating medical image segmentation methods. However, the existing segmentation methods do not attempt to reduce HD directly. In this paper, we present novel loss functions for training convolutional neural network (CNN)-based segmentation methods with the goal of reducing HD directly. We propose three methods to estimate HD from the segmentation probability map produced by a CNN. One method makes use of the distance transform of the segmentation boundary. Another method is based on applying morphological erosion on the difference between the true and estimated segmentation maps. The third method works by applying circular/spherical convolution kernels of different radii on the segmentation probability maps. Based on these three methods for estimating HD, we suggest three loss functions that can be used for training to reduce HD. We use these loss functions to train CNNs for segmentation of the prostate, liver, and pancreas in ultrasound, magnetic resonance, and computed tomography images and compare the results with commonly-used loss functions. Our results show that the proposed loss functions can lead to approximately 18-45% reduction in HD without degrading other segmentation performance criteria such as the Dice similarity coefficient. The proposed loss functions can be used for training medical image segmentation methods in order to reduce the large segmentation errors.

Lightweight image super-resolution with sliding Proxy Attention Network

Summary The interpretation of faults within a geological basin or reservoir from seismic data is a time-consuming, and often manual task associated with high uncertainties. Recently, numerous approaches using machine learning, especially various types of convolutional neural networks, have been presented to automate the process of identifying fault planes within seismic images, which have been shown to outperform traditional fault detection techniques. While these proposed methods show good performance, many of these approaches do not allow investigation of the associated uncertainties that arise in the fault identification process. In this study, we present an application of Bayesian deep convolutional neural networks for identifying faults within seismic datasets. Using an approximate Bayesian inference method a Bayesian deep neural network was trained on a large dataset of synthetic faulted seismic images. The model is then applied to a benchmark dataset and a real data case from NW shelf Australia to identify fault planes, and to investigate the associated uncertainty in the predictive distribution.

Seismic reflection data are used by geologists to identify the best site for oil and gas explorations. The raw seismic image cannot be directly used for explorations because the noises may hinder the primary reflections and result in misinterpretations. In the proposed research work, the denoising method with adaptive dictionary learning is attempted to attenuate the coherent and random noises which affect the seismic reflections. In this method, Karhunen Loeve Transform (KLT) is combined with K-means Singular Value Decomposition (KSVD) to improve the Signal to Noise Ratio (SNR) of the seismic data. The combination of fixed transform using KLT and learning-based dictionary using KSVD remove the redundant data while retaining the necessary data for further interpretations. KLT is applied on the whole seismic image to decorrelate the coefficients and retain the primary reflections. The horizontal events in the seismic image are preserved as they represent the eigen images with large energy. The KSVD is applied to the KLT resultant data, which denoises the patch of data while simultaneously updating the dictionary. The noise is eliminated using local sparsity of the image where mutually overlapping small image blocks are learned to yield the self-adaptive redundant dictionary. It is then used to obtain the sparse representation of the image blocks by eliminating noise. The algorithm is tested for both synthetic and field seismic images, and the results indicate a better reduction of random and coherent noise with the acceptable execution time compared to other denoising algorithms used. The KLT combined with KSVD method decorrelate the seismic data from noise by preserving the primary reflections and discarding the redundant noise.

In contrast to sophisticated means of visual su-per resolution (SR), not much work has been done in the tactile SR field. Existing tactile SR algorithms for taxel-based sensors mainly focus on enhancing the localization accuracy, and generally associate with a specific type of hardware, sometimes not applicable to generic taxel-based tactile sensors. Inspired by image SR, we investigate the tactile pattern SR in this paper, and present how to transform successful image SR schemes, e.g. Convolutional Neural Network (CNN) and Generative Adversarial Network (GAN) to serve the tactile SR. We propose two tactile SR models, i.e. TactileSRCNN and TactileSRGAN, and establish a new tactile pattern SR dataset for model learning. The ground truth of high resolution (HR) tactile patterns in the dataset is obtained via multi-sampling (i.e. overlapping reception) and registration of low resolution (LR) sensor. One key contribution of this research lies in achieving ×100 (from 3×4×4 to 40×40) times tactile pattern SR with a one-time tapping of 3-axis taxel-based sensor. Different from existing tactile SR algorithms which improves the localization accuracy of a single contact point, the proposed scheme can provide multi-point contact detection to robotic applications.

A balanced dataset is generally beneficial to underwater acoustic target recognition. However, the imbalanced class distribution is always meted out in a real scene. To address this, a weighted cross entropy loss function based on trigonometric function is proposed. Then, the proposed loss function is applied in a multi-scale residual convolutional neural network (named MR-CNN-A network) embedded with an attention mechanism for the recognition task. Firstly, a multi-scale convolution kernel is used to obtain multi-scale features. Then, an attention mechanism is used to fuse these multi-scale feature maps. Furthermore, a cosx-function-weighted cross-entropy loss function is used to deal with the class imbalance in underwater acoustic data. This function adjusts the loss ratio of each sample by adjusting the loss interval of every mini-batch based on cosx term to achieve a balanced total loss for each class. Two imbalanced underwater acoustic data sets, ShipsEar and autonomous underwater vehicle (self-collected data) are used to evaluate the proposed network. The experimental results show that the proposed network outperforms the support vector machine and a simple convolutional neural network. Compared with the other three loss functions, the proposed loss function achieves better stability and adaptability. The results strongly demonstrate the validity of the proposed loss function and the network.

Seismic random noise suppression by using deep residual U-Net