Demonstrating multiple attenuation with model-driven processing using neural networks

2021 ◽  
Vol 40 (11) ◽  
pp. 831-836
Author(s):  
Aina Juell Bugge ◽  
Andreas K. Evensen ◽  
Jan Erik Lie ◽  
Espen H. Nilsen
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Random Noise ◽  
Synthetic Data ◽  
Training Data ◽  
Data Set ◽  
Amplitude Variation With Offset ◽  
Model Driven ◽  
Multiple Attenuation ◽  
Common Depth Point

Some of the key tasks in seismic processing involve suppressing multiples and noise that interfere with primary events. Conventional multiple attenuation on seismic prestack data is time-consuming and subjective. As an alternative, we propose model-driven processing using a convolutional neural network trained on synthetically modeled training data. The crucial part of our approach is to generate appropriate training data. Here, we compute a generic data set with pairs of synthetic gathers with and without multiples. Because we generate the primaries first and then add multiples, we ensure that we have perfect target data without any multiple energy. To compute generic and realistic training data, we include elements of wave propagation physics and implement a randomized flexibility of settings such as the wavelet, frequency content, degree of random noise, and amplitude variation with offset effects with each gather pair. A fully convolutional neural network is trained on the synthetic data in order to learn to suppress the noise and multiples. Evaluations of the approach on benchmark data indicate that our trained network is faster than conventional multiple attenuation because it can be run efficiently on a modern GPU, and it has the potential to better preserve primary amplitudes. Multiple removal with model-driven processing is demonstrated on seismic field data, and the results are compared to conventional multiple attenuation using a commercial Radon algorithm. The model-driven approach performs well when applied to real common-depth point gathers, and it successfully removes multiples, even where the multiples interfere with the primary signals on the near offsets.

Automatic velocity analysis using convolutional neural network and transfer learning

Geophysics ◽  
2019 ◽  
Vol 85 (1) ◽  
pp. V33-V43 ◽  
Author(s):  
Min Jun Park ◽  
Mauricio D. Sacchi
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Transfer Learning ◽  
Input Data ◽  
Synthetic Data ◽  
Velocity Model ◽  
Training Data ◽  
Velocity Analysis ◽  
Time Step ◽  
Data Set

Velocity analysis can be a time-consuming task when performed manually. Methods have been proposed to automate the process of velocity analysis, which, however, typically requires significant manual effort. We have developed a convolutional neural network (CNN) to estimate stacking velocities directly from the semblance. Our CNN model uses two images as one input data for training. One is an entire semblance (guide image), and the other is a small patch (target image) extracted from the semblance at a specific time step. Labels for each input data set are the root mean square velocities. We generate the training data set using synthetic data. After training the CNN model with synthetic data, we test the trained model with another synthetic data that were not used in the training step. The results indicate that the model can predict a consistent velocity model. We also noticed that when the input data are extremely different from those used for the training, the CNN model will hardly pick the correct velocities. In this case, we adopt transfer learning to update the trained model (base model) with a small portion of the target data to improve the accuracy of the predicted velocity model. A marine data set from the Gulf of Mexico is used for validating our new model. The updated model performed a reasonable velocity analysis in seconds.

scholarly journals Convolutional Neural Network-Based Discriminator for Outlier Detection

2021 ◽  
Vol 2021 ◽  
pp. 1-13
Author(s):  
Fahad Alharbi ◽  
Khalil El Hindi ◽  
Saad Al Ahmadi ◽  
Hussien Alsalamn
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Systematic Approach ◽  
Data Augmentation ◽  
Random Noise ◽  
Training Data ◽  
Human Errors ◽  
Machine Learning Methods ◽  
Benchmark Datasets ◽  
Trusted Data

Noise in training data increases the tendency of many machine learning methods to overfit the training data, which undermines the performance. Outliers occur in big data as a result of various factors, including human errors. In this work, we present a novel discriminator model for the identification of outliers in the training data. We propose a systematic approach for creating training datasets to train the discriminator based on a small number of genuine instances (trusted data). The noise discriminator is a convolutional neural network (CNN). We evaluate the discriminator’s performance using several benchmark datasets and with different noise ratios. We inserted random noise in each dataset and trained discriminators to clean them. Different discriminators were trained using different numbers of genuine instances with and without data augmentation. We compare the performance of the proposed noise-discriminator method with seven other methods proposed in the literature using several benchmark datasets. Our empirical results indicate that the proposed method is very competitive to the other methods. It actually outperforms them for pair noise.

scholarly journals Skin Cancer Detection using CNN Algorithm

2020 ◽  
Vol 9 (6) ◽  
pp. 45-49
Keyword(s):  
Neural Network ◽  
Skin Cancer ◽  
Convolutional Neural Network ◽  
Well Being ◽  
Training Data ◽  
Data Sets ◽  
Data Set ◽  
Sequential Model ◽  
Pigmented Lesions ◽  
Testing Data

The project “Disease Prediction Model” focuses on predicting the type of skin cancer. It deals with constructing a Convolutional Neural Network(CNN) sequential model in order to find the type of a skin cancer which takes a huge troll on mankind well-being. Since development of programmed methods increases the accuracy at high scale for identifying the type of skin cancer, we use Convolutional Neural Network, CNN algorithm in order to build our model . For this we make use of a sequential model. The data set that we have considered for this project is collected from NCBI, which is well known as HAM10000 dataset, it consists of massive amounts of information regarding several dermatoscopic images of most trivial pigmented lesions of skin which are collected from different sufferers. Once the dataset is collected, cleaned, it is split into training and testing data sets. We used CNN to build our model and using the training data we trained the model , later using the testing data we tested the model. Once the model is implemented over the testing data, plots are made in order to analyze the relation between the echos and loss function. It is also used to analyse accuracy and echos for both training and testing data.

scholarly journals RainNet v1.0: a convolutional neural network for radar-based precipitation nowcasting

2020 ◽  
Vol 13 (6) ◽  
pp. 2631-2644 ◽  
Author(s):  
Georgy Ayzel ◽  
Tobias Scheffer ◽  
Maik Heistermann
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Lead Time ◽  
Spectral Power ◽  
Training Data ◽  
Lead Times ◽  
Continuous Precipitation ◽  
Data Set ◽  
Binary Segmentation ◽  
Model Training

Abstract. In this study, we present RainNet, a deep convolutional neural network for radar-based precipitation nowcasting. Its design was inspired by the U-Net and SegNet families of deep learning models, which were originally designed for binary segmentation tasks. RainNet was trained to predict continuous precipitation intensities at a lead time of 5 min, using several years of quality-controlled weather radar composites provided by the German Weather Service (DWD). That data set covers Germany with a spatial domain of 900 km×900 km and has a resolution of 1 km in space and 5 min in time. Independent verification experiments were carried out on 11 summer precipitation events from 2016 to 2017. In order to achieve a lead time of 1 h, a recursive approach was implemented by using RainNet predictions at 5 min lead times as model inputs for longer lead times. In the verification experiments, trivial Eulerian persistence and a conventional model based on optical flow served as benchmarks. The latter is available in the rainymotion library and had previously been shown to outperform DWD's operational nowcasting model for the same set of verification events. RainNet significantly outperforms the benchmark models at all lead times up to 60 min for the routine verification metrics mean absolute error (MAE) and the critical success index (CSI) at intensity thresholds of 0.125, 1, and 5 mm h−1. However, rainymotion turned out to be superior in predicting the exceedance of higher intensity thresholds (here 10 and 15 mm h−1). The limited ability of RainNet to predict heavy rainfall intensities is an undesirable property which we attribute to a high level of spatial smoothing introduced by the model. At a lead time of 5 min, an analysis of power spectral density confirmed a significant loss of spectral power at length scales of 16 km and below. Obviously, RainNet had learned an optimal level of smoothing to produce a nowcast at 5 min lead time. In that sense, the loss of spectral power at small scales is informative, too, as it reflects the limits of predictability as a function of spatial scale. Beyond the lead time of 5 min, however, the increasing level of smoothing is a mere artifact – an analogue to numerical diffusion – that is not a property of RainNet itself but of its recursive application. In the context of early warning, the smoothing is particularly unfavorable since pronounced features of intense precipitation tend to get lost over longer lead times. Hence, we propose several options to address this issue in prospective research, including an adjustment of the loss function for model training, model training for longer lead times, and the prediction of threshold exceedance in terms of a binary segmentation task. Furthermore, we suggest additional input data that could help to better identify situations with imminent precipitation dynamics. The model code, pretrained weights, and training data are provided in open repositories as an input for such future studies.

scholarly journals Application of a convolutional neural network for seismic phase picking of mining-induced seismicity

2020 ◽  
Vol 224 (1) ◽  
pp. 230-240
Author(s):  
Sean W Johnson ◽  
Derrick J A Chambers ◽  
Michael S Boltz ◽  
Keith D Koper
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Induced Seismicity ◽  
Southern California ◽  
High Stress ◽  
Training Data ◽  
Data Sets ◽  
Residual Variance ◽  
Data Set ◽  
Mining Induced Seismicity

SUMMARY Monitoring mining-induced seismicity (MIS) can help engineers understand the rock mass response to resource extraction. With a thorough understanding of ongoing geomechanical processes, engineers can operate mines, especially those mines with the propensity for rockbursting, more safely and efficiently. Unfortunately, processing MIS data usually requires significant effort from human analysts, which can result in substantial costs and time commitments. The problem is exacerbated for operations that produce copious amounts of MIS, such as mines with high-stress and/or extraction ratios. Recently, deep learning methods have shown the ability to significantly improve the quality of automated arrival-time picking on earthquake data recorded by regional seismic networks. However, relatively little has been published on applying these techniques to MIS. In this study, we compare the performance of a convolutional neural network (CNN) originally trained to pick arrival times on the Southern California Seismic Network (SCSN) to that of human analysts on coal-mine-related MIS. We perform comparisons on several coal-related MIS data sets recorded at various network scales, sampling rates and mines. We find that the Southern-California-trained CNN does not perform well on any of our data sets without retraining. However, applying the concept of transfer learning, we retrain the SCSN model with relatively little MIS data after which the CNN performs nearly as well as a human analyst. When retrained with data from a single analyst, the analyst-CNN pick time residual variance is lower than the variance observed between human analysts. We also compare the retrained CNN to a simpler, optimized picking algorithm, which falls short of the CNN's performance. We conclude that CNNs can achieve a significant improvement in automated phase picking although some data set-specific training will usually be required. Moreover, initializing training with weights found from other, even very different, data sets can greatly reduce the amount of training data required to achieve a given performance threshold.

scholarly journals Synthetic Sensor Array Training Sets for Neural Networks

2019 ◽  
Vol 2019 ◽  
pp. 1-10 ◽  
Author(s):  
Oded Medina ◽  
Roi Yozevitch ◽  
Nir Shvalb
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Sensor Array ◽  
Synthetic Data ◽  
Training Data ◽  
Small Data ◽  
Data Set ◽  
Training Time ◽  
Wide Range ◽  
Artificial Neural

It is often hard to relate the sensor’s electrical output to the physical scenario when a multidimensional measurement is of interest. An artificial neural network may be a solution. Nevertheless, if the training data set is extracted from a real experimental setup, it can become unreachable in terms of time resources. The same issue arises when the physical measurement is expected to extend across a wide range of values. This paper presents a novel method for overcoming the long training time in a physical experiment set up by bootstrapping a relatively small data set for generating a synthetic data set which can be used for training an artificial neural network. Such a method can be applied to various measurement systems that yield sensor output which combines simultaneous occurrences or wide-range values of physical phenomena of interest. We discuss to which systems our method may be applied. We exemplify our results on three study cases: a seismic sensor array, a linear array of strain gauges, and an optical sensor array. We present the experimental process, its results, and the resulting accuracies.

Prestack and poststack inversion using a physics-guided convolutional neural network

2019 ◽  
Vol 7 (3) ◽  
pp. SE161-SE174 ◽  
Author(s):  
Reetam Biswas ◽  
Mrinal K. Sen ◽  
Vishal Das ◽  
Tapan Mukerji
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Elastic Properties ◽  
Wave Velocity ◽  
Seismic Data ◽  
Synthetic Data ◽  
P Wave ◽  
Distinct Advantage ◽  
Elastic Parameters ◽  
Data Set

An inversion algorithm is commonly used to estimate the elastic properties, such as P-wave velocity ([Formula: see text]), S-wave velocity ([Formula: see text]), and density ([Formula: see text]) of the earth’s subsurface. Generally, the seismic inversion problem is solved using one of the traditional optimization algorithms. These algorithms start with a given model and update the model at each iteration, following a physics-based rule. The algorithm is applied at each common depth point (CDP) independently to estimate the elastic parameters. Here, we have developed a technique using the convolutional neural network (CNN) to solve the same problem. We perform two critical steps to take advantage of the generalization capability of CNN and the physics to generate synthetic data for a meaningful representation of the subsurface. First, rather than using CNN as in a classification type of problem, which is the standard approach, we modified the CNN to solve a regression problem to estimate the elastic properties. Second, again unlike the conventional CNN, which is trained by supervised learning with predetermined label (elastic parameter) values, we use the physics of our forward problem to train the weights. There are two parts of the network: The first is the convolution network, which takes the input as seismic data to predict the elastic parameters, which is the desired intermediate result. In the second part of the network, we use wave-propagation physics and we use the output of the CNN to generate the predicted seismic data for comparison with the actual data and calculation of the error. This error between the true and predicted seismograms is then used to calculate gradients, and update the weights in the CNN. After the network is trained, only the first part of the network can be used to estimate elastic properties at remaining CDPs directly. We determine the application of physics-guided CNN on prestack and poststack inversion problems. To explain how the algorithm works, we examine it using a conventional CNN workflow without any physics guidance. We first implement the algorithm on a synthetic data set for prestack and poststack data and then apply it to a real data set from the Cana field. In all the training examples, we use a maximum of 20% of data. Our approach offers a distinct advantage over a conventional machine-learning approach in that we circumvent the need for labeled data sets for training.

scholarly journals Multi-attribute self-attention guided vehicle local region detection based on convolutional neural network architecture

2020 ◽  
Vol 17 (4) ◽  
pp. 172988142094434
Author(s):  
Jingbo Chen ◽  
Shengyong Chen ◽  
Linjie Bian
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Quantitative Evaluation ◽  
Network Architecture ◽  
Local Region ◽  
Training Data ◽  
Detection Methods ◽  
Data Set ◽  
Region Detection ◽  
Bounding Boxes

Many pieces of information are included in the front region of a vehicle, especially in windshield and bumper regions. Thus, windshield or bumper region detection is making sense to extract useful information. But the existing windshield and bumper detection methods based on traditional artificial features are not robust enough. Those features may become invalid in many real situations (e.g. occlude, illumination change, viewpoint change.). In this article, we propose a multi-attribute-guided vehicle discriminately region detection method based on convolutional neural network and not rely on bounding box regression. We separate the net into two branches, respectively, for identification (ID) and Model attributes training. Therefore, the feature spaces of different attributes become more independent. Additionally, we embed a self-attention block into our framework to improve the performance of local region detection. We train our model on PKU_VD data set which has a huge number of images inside. Furthermore, we labeled the handcrafted bounding boxes on 5000 randomly picked testing images, and 1020 of them are used for evaluation and 3980 as the training data for YOLOv3. We use Intersection over Union for quantitative evaluation. Experiments were conducted in three different latest convolutional neural network trunks to illustrate the detection performance of the proposed method. Simultaneously, in terms of quantitative evaluation, the performance of our method is close to YOLOv3 even without handcrafted bounding boxes.

scholarly journals RainNet v1.0: a convolutional neural network for radar-based precipitation nowcasting

2020 ◽  
Author(s):  
Georgy Ayzel ◽  
Tobias Scheffer ◽  
Maik Heistermann
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Lead Time ◽  
Spectral Power ◽  
Training Data ◽  
Lead Times ◽  
Continuous Precipitation ◽  
Data Set ◽  
Binary Segmentation ◽  
Model Training

Abstract. In this study, we present RainNet, a deep convolutional neural network for radar-based precipitation nowcasting. Its design was inspired by the U-Net and SegNet families of deep learning models which were originally designed for binary segmentation tasks. RainNet was trained to predict continuous precipitation intensities at a lead time of five minutes, using several years of quality-controlled weather radar composites provided by the German Weather Service (DWD). That data set covers Germany with a spatial domain of 900 by 900 km, and has a resolution of 1 km in space and 5 minutes in time. Independent verification experiments were carried out on eleven summer precipitation events from 2016 to 2017. In order to achieve a lead time of one hour, a recursive approach was implemented by using RainNet predictions at five minutes lead time as model input for longer lead times. In the verification experiments, trivial Eulerian persistence and a conventional model based on optical flow served as benchmarks. The latter is available in the rainymotion library, and had previously been shown to outperform DWD's operational nowcasting model for the same set of verification events. RainNet significantly outperforms the benchmark models at all lead times up to 60 minutes for the routine verification metrics Mean Absolute Error (MAE) and Critical Success Index (CSI, at intensity thresholds of 0.125, 1, and 5 mm/h). Apart from its superiority in terms of MAE and CSI, an undesirable property of RainNet predictions is, however, the level of spatial smoothing. At a lead time of five minutes, an analysis of Power Spectral Density confirmed a significant loss of spectral power at length scales of 16 km and below. Obviously, RainNet had learned an optimal level of smoothing to produce a nowcast at 5 minutes lead time. In that sense, the loss of spectral power at small scales is informative, too, as it reflects the limits of predictability as a function of spatial scale. Beyond the lead time of five minutes, however, the increasing level of smoothing is a mere artifact -- an analogue to numerical diffusion -- that is not a property of RainNet itself, but of its recursive application. In the context of early warning, the smoothing is particularly unfavourable since pronounced features of intense precipitation tend to get lost over longer lead times. Hence, we propose several options to address this issue in prospective research, including an adjustment of the loss function for model training, model training for longer lead times, and the prediction of threshold exceedance in terms of a binary segmentation task. Furthermore, we suggest additional input data that could help to better identify situations with imminent precipitation dynamics. The model code, pretrained weights, and training data are provided in open repositories as an input to such future studies.

Sign in / Sign up

Export Citation Format

Share Document