... neural network machine learning architecture american association alignment procedure seismic amplitude boundary augmentation exploration geophysicist applied geoscience energy 10 society compute outlier rectangle An integrated workflow of improving the accuracy of first arrivals...

... We propose a method to generate seismic images with corresponding fault labels for augmenting training data in automatic faults detection. We adopt two generative adversarial networks: one network for creating faults system and the other for generating 2D seismic images with faults...

... recognition workflow ssl iteration exploration geophysicist 10 augmentation baseline model salt interpretation artificial intelligence applied geoscience information unlabeled data ground truth subsurface salt recognition Subsurface salt recognition via deep learning- an iterative semi...

... training data, Fancy implementing the mineral inversion requires the exact principal component analysis was adopted to augment theoretical response value of each mineral. Therefore, in this training data before training the DNN model. Blind test was paper, we propose a deep neural network (DNN), which does...

... Presentation Time: 10:35 AM Location: Poster Station 9 Presentation Type: Poster reservoir characterization upstream oil & gas training set machine learning deep learning data augmentation artificial intelligence classification generator neural network validation set training data seg...

... Date: Wednesday, September 18, 2019 Session Start Time: 1:50 PM Presentation Time: 3:05 PM Location: Poster Station 2 Presentation Type: Poster test image machine learning loss function salt body interpretation detection segmentation neural network deep learning augmentation deep...

... ABSTRACT The low frequency component is essential for inverting seismic reflection data. Various methods have been proposed to augment the low frequencies. However, one common limitation for these methods is that the data gap (missing low frequencies) can’t be large compared to the Nyquist...

... Summary We have developed a method to augment reflection data in the missing low-frequency range for acoustic Inverse Scattering Series. It relies on a least square fitting using Hermite Distributed Approximation Functionals (HDAF). This particular choice is based on the property that DAF has...