...-limited data as pseudo labels, with input data derived by filtering out highfrequency elements from the original data. Subsequently, the network undergoes iterative data refinement, where pseudo labels are predicted from the NN trained in the previous epoch on the original data, and input data...

... separately or vice versa. We show both synthetic and real data examples to demonstrate that our proposed method is a good alternative for input data with mixed curvatures. geologist curvature reconstruction side show reservoir characterization radon transformation iteration forward...

... geology reservoir characterization artificial intelligence lstm deconvolution dataset seismic data bi-directional lstm-based applied geoscience & energy 10 spiking deconvolution input data interpretation frequency deep learning machine learning spectrum exploration geophysicist...

..., 1986). TTI-RTM is required for seismic imaging. Seismic data acquired in these u( , p) ds n (x,t px) (1) environments is also distorted by the rugged x topography and near surface fluvial deposit. Strong direct wave, guided wave, refracted wave and Where, ds n (x, t) is the input data after...

... run we simulate data with surface decoder parts, where the encoder projects input data into latent multiples by applying free-surface boundary condition at the space, Rh×wà Rc×n×m, while the decoder is trained to do the surface and absorbing boundaries at other edges. In the second reverse mapping...

... reservoir characterization tlfwi kirchhoff image velocity detail upstream oil & gas orphan basin high-order multiple exploration geophysicist 10 anomaly expanded abstract velocity model applied geoscience prediction input data order multiple limitation fwi imaging seismic...

... have been developed to attenuate internal mul- internal multiple. For an incident spike wave, b1c0) = tiples. Among these methods, Delft s feedback method (Berkhout D which is the input data. b1(z) corresponds to an uncol- and Verschuur, 1997) and Jackubowicz s convolution-correlation lapsed FK...

... Self-organizing mapping (SOM) is one of the most famous classification method in seismic facies analysis. Traditional self-organizing map network and its variation methods usually use the Euclidean distance to measure the similarity between input data and weights of neuron node. A Euclidean...

.... Tomographic waveform inversion is designed to handle complex geology with highly heterogeneous P-wave velocity and at same time overcome the typical challenges of other waveform inversion techniques. upstream oil & gas subsurface input data operator equation reservoir characterization applied...

... is frequency domain, has been proposed to attenuate ground to decompose the input data into eigenimages with SVD, roll. The key concept is to construct a noise model from where a few eigenimages that correspond to the largest eigenimages and remove it from input data. The main task is eigenvalues are selected...

... when compared with conventional 4D approaches. applied geoscience input data migration difference noise reservoir characterization repeatability velocity difference fwi joint time-lapse full-waveform inversion field data example perfect repeatability velocity model exploration...

..., and near-bit vibrations. We used a machine learning-assisted workflow to model the wireline logs and assess the importance of various input data. The use of nearbit vibrations reduces the prediction error (of sonic data?) from ∼5.3% achieved relying on to 1.8%, which is sufficient for drilling/completion...

... impossible to achieve by other seismic imaging approaches. resolution lateral resolution input data fwi imaging frequency fwi image kirchhoff image hz tlfwi model upstream oil & gas applied geoscience expanded abstract inversion reservoir characterization hz fwi image seismic...

... Session Start Time: 1:50 PM Presentation Time: 3:55 PM Location: Poster Station 13 Presentation Type: Poster artificial intelligence convolution input data neural network field data exploration geophysicist 10 demultiple alexander breuer unified approach machine learning deep learning...

... result algorithm seismic data well-log data seg international exposition machine learning deep learning estimate elastic parameter input data estimation parameter estimation elastic parameter estimation Elastic parameters estimation using multi-model based on machine learning Kai Xu...

...: Poster Station 9 Presentation Type: Poster reservoir characterization p-wave impedance input data seg international exposition upstream oil & gas elastic impedance exploration geophysicist 10 algorithm machine learning objective function stability stable rock-physic inversion...

... passive seismic the central trace to allow the network to focus on the curvature events that are recorded by two stations to roughly estimate the of the input data. We train the network to handle both sin- source locations (Perol et al., 2018). Kriegerowski et al. (2018) gle and double events that might...

... the network with as input data and neural network and predict the final low-frequency data, we as target data. apply another high-cut filter to mimic the training scenario but create inputs for prediction. These processes will Assume the pattern learned from the training workflow can eventually help generate...

... data show that the algorithm is very effective. Although not promotes sparsity in Fm while the equality constraint ensures presented in this abstract, we experience equally effective re- that the re-blended solution matches the input data. Of the many methods available, we use an iterative thresholding...

... We propose a multi-channel dynamic matching fullwaveform inversion (DMFWI) for a high-resolution velocity-model update, which focuses on solving kinematic difference between input data and synthetic data. With the data residual calculated in localized windows in time and space, DMFWI provides...