Not enough data to create a plot.
Try a different view from the menu above.
Image Processing
Introduction to special section: Borehole image data applications in reservoir characterization ? Case studies and updates on new developments
Li, Bingjian (Blackriver Geoscience LLC) | Egorov, Vsevolod (GeoExpera) | Perona, Ricardo (Repsol) | Haddad, David (Apache Corporation) | Sementelli, Katy (Woodside Energy) | Xu, Chicheng (Aramco Americas Company) | Mardi, Chrystianto (BPX Energy)
Borehole image data have played an important role in the oil and gas industry for decades, providing invaluable insights into hydrocarbon exploration, reservoir appraisal, and development. Recent advancements in borehole image technologies, encompassing data acquisition, processing, and interpretation, have ushered in a new era of possibilities. Geoscientists have expanded the applications of image data, progressing from basic natural fracture detection to comprehensive reservoir characterization. This special section explores significant advances in sedimentological and structural interpretation, full-scale fracture and fault analysis, wellbore geomechanics, reservoir heterogeneity evaluation, and 3-D reservoir modeling. Applications of borehole image log data have transcended reservoir types, spanning clastics, carbonates, naturally fractured basements, and unconventional shales. With these developments in mind, we have invited submissions that showcase studies utilizing borehole image log data for the successful characterization of any reservoir type, along with related case studies of interest to the exploration and development community. The overwhelming response to our call-for-papers resulted in the selection of seven high-quality contributions for inclusion in this special publication. Mohebian et al. revolutionize fracture identification by employing the YOLOv5 deep learning algorithm on borehole image logs, shifting from manual to automated processes.
- Geology > Geological Subdiscipline > Geomechanics (1.00)
- Geology > Rock Type > Sedimentary Rock > Clastic Rock (0.56)
- Geophysics > Seismic Surveying > Borehole Seismic Surveying (1.00)
- Geophysics > Borehole Geophysics (1.00)
- Asia > Kazakhstan > West Kazakhstan > Uralsk Region > Precaspian Basin > Karachaganak Field (0.99)
- Asia > China > Bohai Basin (0.99)
- Information Technology > Sensing and Signal Processing > Image Processing (0.86)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks > Deep Learning (0.58)
LITHOCODIUM MOUND IDENTIFICATION USING LWD IMAGE LOG AND QUANTIFIED CUTTING ANALYSIS ย VALIDATION WITH ANALOGUES
Perrin, Christian (North Oil Company) | Pointer, Chay (North Oil Company) | Al-Mohannadi, Ghada (North Oil Company) | Sen, Shantanu (North Oil Company) | Buraimoh, Muse Ajadi (QatarEnergy)
Lithocodium mounds are early Cretaceous sedimentary structures described in the literature from outcrops, however, never described in the subsurface. The objective of this work is to identify and characterize Lithocodium mounds in the subsurface along a 25,000ft horizontal well. Drill cuttings sampled at a 100ft interval are observed in thin sections to define and quantify key sedimentary indicators (bioclasts, facies, and texture). Logging-while-drilling (LWD) GR, density, neutron, and resistivity logs are acquired along with the LWD high-resolution borehole image (BHI) log. Bedding dips from BHI data, interpreted along the horizontal well, enabled the reconstruction of the reservoir paleotopography. In particular, the alternation of dip azimuth combined with the facies interpretation from the thin sections supported the interpretation of eight distinct mound structures. An assessment of their overall geometry confirmed the mound shape to be subcircular, consistent with the subcircular geometries observed in Oman at the outcrop. The inferred dimensions of the mounds are comparable with the Aptian Lithocodium mounds in Oman (3040m), and their intermound organization resembles that of the Albian mounds in Texas. This work demonstrates the value of analyzing cuttings to complement image log interpretation and the value of outcrop analogs for interpreting sedimentary structures. For the first time, the subsurface identification and characterization of Lithocodium mounds and intermounds are achieved.
- North America > United States > Texas (0.48)
- Asia > Middle East > Oman (0.45)
- Geology > Rock Type > Sedimentary Rock > Carbonate Rock (1.00)
- Geology > Sedimentary Geology > Depositional Environment (0.93)
- Geology > Geological Subdiscipline > Stratigraphy (0.66)
- Geophysics > Borehole Geophysics (1.00)
- Geophysics > Seismic Surveying > Borehole Seismic Surveying (0.48)
- Well Drilling > Drilling Operations (1.00)
- Well Drilling > Drilling Measurement, Data Acquisition and Automation > Logging while drilling (1.00)
- Reservoir Description and Dynamics > Reservoir Characterization (1.00)
- Reservoir Description and Dynamics > Formation Evaluation & Management > Open hole/cased hole log analysis (1.00)
Quantitative characterization of organic and inorganic pores in shale based on deep learning
Yan, Bohong (China University of Petroleum) | Sun, Langqiu (China University of Petroleum) | Zhao, Jianguo (China University of Petroleum) | Cao, Zixiong (Object Research Systems (ORS) Company) | Li, Mingxuan (China University of Petroleum) | Shiba, K. C. (China University of Petroleum) | Liu, Xinze (Yumen Oil Field Branch of China National Petroleum Corporation (CNPC) Exploration and Development Research Institute) | Li, Chuang (China National Petroleum Corporation (CNPC))
ABSTRACT Organic matter (OM) maturity is closely related to organic pores in shales. Quantitative characterization of organic and inorganic pores in shale is crucial for rock-physics modeling and reservoir porosity and permeability evaluation. Focused ion beam-scanning electron microscopy (FIB-SEM) can capture high-precision three-dimensional (3D) images and directly describe the types, shapes, and spatial distribution of pores in shale gas reservoirs. However, due to the high scanning cost, wide 3D view field, and complex microstructure of FIB-SEM, more efficient segmentation for the FIB-SEM images is required. For this purpose, a multiphase segmentation workflow in conjunction with a U-net is developed to segment pores from the matrix and distinguish organic pores from inorganic pores simultaneously in the entire 3D image stack. The workflow is repeated for FIB-SEM data sets of 17 organic-rich shales with various characteristics. The analysis focuses on improving the efficiency and relevance of the workflow, that is, quantifying the minimum number of training slices while ensuring accuracy and further combining the fractal dimension (FD) and lacunarity to study a simple and objective method of selection. Meanwhile, the computational efficiency, accuracy, and robustness to noise of the 2D U-net model are discussed. The intersection over the union of automatic segmentation can amount to 80%โ95% in all data sets with manual labels as ground truth. In addition, calculated by the FIB-SEM multiphase segmentation, the organic porosity is used to quantitatively evaluate the OM decomposition level. Deep-learning-based segmentation shows great potential for characterizing shale pore structures and quantifying OM maturity.
- Asia > China (1.00)
- North America > United States > Texas (0.68)
- North America > United States > Texas > West Gulf Coast Tertiary Basin > Eagle Ford Shale Formation (0.99)
- North America > United States > Texas > Sabinas - Rio Grande Basin > Eagle Ford Shale Formation (0.99)
- North America > United States > Texas > Maverick Basin > Eagle Ford Shale Formation (0.99)
- (7 more...)
- Reservoir Description and Dynamics > Unconventional and Complex Reservoirs > Shale gas (1.00)
- Reservoir Description and Dynamics > Reservoir Characterization > Exploration, development, structural geology (1.00)
- Data Science & Engineering Analytics > Information Management and Systems > Neural networks (1.00)
- Information Technology > Sensing and Signal Processing > Image Processing (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks > Deep Learning (1.00)
Use of Resistivity and Density Borehole Image Logs to Identify and Distribute Facies in the Pikka Unit - A Case Study from the Nanushuk Formation, North Slope, Alaska
Perona, Ricardo (Repsol USA) | Armitage, Dominic (Repsol USA) | Bonelli, James (Repsol USA) | Capuzzo, Nicola (Task Fronterra Geoscience) | Tingey, Brady (Task Fronterra Geoscience)
Over the past decade, the North Slope of Alaska has yielded several major hydrocarbon discoveries in deltaic topsets of the Brookian Nanushuk Formation. Together the Nanushuk topsets and genetically related foreset and bottomset beds of the Torok Formation comprise part of a giant clinothem system that prograded across the Colville Foreland Basin during the lower Cretaceous (Aptian through Cenomanian). The discovered Nanushuk topset play contains stratigraphically trapped hydrocarbons within multiple fairways trending roughly north to south along the basinยs extent. The Nanushuk topset play was first discovered in the Pikka Unit by Repsol and partners during the 2013 winter drilling campaign. The Pikka Unit is located at the eastern edge of the Nanushuk-Torok clinothem system and underlies the modern-day Colville River. Here, the Nanushuk Formation comprises shelf-edge deltaic and shoreface deposits, characterized by intercalations of fine-grained litharenites and silty mudstones. The layered character of the formation is readily recognized in electric logs due to density and resistivity contrasts between those main lithologies. Following the initial Pikka discovery, 14 appraisal wells were drilled in the unit, including 2 horizontal and 2 high angle wells. An extensive and diverse borehole image dataset was acquired and includes wireline high resolution oil-based mud resistivity and logging-while-drilling azimuthal density images. In addition, more than 1000 feet of continuous core was collected in three wells (Qugruk-8, Pikka B, Pikka B ST1). Borehole images were then used to orientate the high resolution CT scan images of the cores, which afterwards were integrated with the image log analysis. This study presents a case on how the integration of core sedimentology and detailed borehole image log analysis were used to guide and predict the facies distribution across the Pikka unit.
- Geology > Geological Subdiscipline > Stratigraphy (1.00)
- Geology > Rock Type > Sedimentary Rock > Clastic Rock > Mudrock (0.68)
- Geology > Sedimentary Geology > Depositional Environment > Transitional Environment > Deltaic Environment (0.54)
- Geophysics > Seismic Surveying > Borehole Seismic Surveying (1.00)
- Geophysics > Borehole Geophysics (1.00)
- North America > United States > Alaska > North Slope Basin > Umiat-Gubik Area > Torok Formation (0.99)
- North America > United States > Alaska > North Slope Basin > Prudhoe Bay Field (0.99)
- North America > United States > Alaska > North Slope Basin > Pikka Unit > Nanushuk Formation (0.99)
- (3 more...)
- Reservoir Description and Dynamics > Reservoir Characterization > Geologic modeling (1.00)
- Reservoir Description and Dynamics > Formation Evaluation & Management > Open hole/cased hole log analysis (1.00)
- Production and Well Operations > Well & Reservoir Surveillance and Monitoring > Borehole imaging and wellbore seismic (1.00)
The main objective of this work is to understand the impact of fracture, stress, drilling direction and other reservoir properties on the production performance in horizontal well (HW). Taking advantage of seventy available borehole image logs helped to extend analysis beyond individual wells to a field scale evaluation. Three analysis techniques were developed to progress with the study: Digital Interpretation of Borehole Breakout in image log, Favored Drilling Direction Map, and a Reservoir Property Filter to gauge well performance. Results in cross plots showed complicated, cloudy and multi-dimensional relationships. The findings will be used to guide future HW drilling optimization, support dynamic modeling and improve modelยs predictability for effective reservoir management.
- North America > United States (0.46)
- Asia > Kazakhstan > West Kazakhstan Region (0.29)
- Phanerozoic > Paleozoic > Permian (0.94)
- Phanerozoic > Paleozoic > Devonian (0.68)
- Geology > Structural Geology > Tectonics (1.00)
- Geology > Geological Subdiscipline > Geomechanics (1.00)
- Geology > Rock Type > Sedimentary Rock (0.93)
- (2 more...)
- Geophysics > Borehole Geophysics (1.00)
- Geophysics > Seismic Surveying > Borehole Seismic Surveying (0.61)
- North America > United States > California > San Joaquin Basin > Lost Hills Field (0.99)
- North America > United States > California > Monterey Formation (0.99)
- Asia > Kazakhstan > West Kazakhstan > Uralsk Region > Precaspian Basin > Karachaganak Field (0.99)
- (5 more...)
- Reservoir Description and Dynamics > Reservoir Characterization > Exploration, development, structural geology (1.00)
- Production and Well Operations > Well & Reservoir Surveillance and Monitoring (1.00)
- Reservoir Description and Dynamics > Unconventional and Complex Reservoirs > Naturally-fractured reservoirs (0.93)
- (5 more...)
In this study, we interpreted a cumulative 600m acoustic image log across the Triassic to Cambro-Ordovician interval in the Berkaoui oil field, Algeria. We interpreted 40 distinct breakout zones which have a combined length of 210m. These breakouts are aligned in the NNE-SSW direction indicating a mean maximum horizontal stress (SHmax) azimuth of 110ยฐN. The observed breakouts are ranked as ยA-Qualityย following the World Stress Map ranking guidelines. The angular width of each breakout has been inferred from the image log analysis and the same has been utilized to infer the SHmax gradient by stress polygon approach following the frictional faulting mechanism. The stress polygon across all the breakout intervals provides a practical Shmax range between 24.7-31.1 MPa/km, with an average gradient of ~ 27 MPa/km. Considering the Shmin range across the studied intervals, we infer a SHmax/Shmin ratio dominantly between 1.40-1.65, which is a much narrower and better-constrained range when compared to the previously published ranges from nearby fields with the same stratigraphy. The relative magnitudes of the in-situ stresses indicate a strike-slip faulting regime in the Berkaoui field. This study presents the utility of image log analysis and integration of breakout interpretation to obtain a more robust geomechanical model with reduced SHmax uncertainty.
- North America > United States (1.00)
- Asia (1.00)
- Africa > Middle East > Algeria > Illizi Province (0.28)
- Africa > Middle East > Algeria > Ouargla Province > Hassi Messaoud (0.28)
- Phanerozoic > Paleozoic > Ordovician (1.00)
- Phanerozoic > Mesozoic > Cretaceous > Upper Cretaceous (0.46)
- Geology > Geological Subdiscipline > Geomechanics (1.00)
- Geology > Geological Subdiscipline > Economic Geology > Petroleum Geology (0.75)
- Geology > Rock Type > Sedimentary Rock > Clastic Rock (0.47)
- Geology > Structural Geology > Tectonics > Plate Tectonics (0.46)
- Geophysics > Borehole Geophysics (1.00)
- Geophysics > Seismic Surveying > Borehole Seismic Surveying (0.72)
- Asia > Middle East > Iraq > Basra Governorate > Arabian Basin > Widyan Basin > Mesopotamian Basin > Zubair Field > Zubair Formation (0.99)
- Asia > Middle East > Iraq > Basra Governorate > Arabian Basin > Widyan Basin > Mesopotamian Basin > Zubair Field > Mishrif Formation (0.99)
- Africa > Middle East > Egypt > Western Desert > Greater Western Dester Basin > Abu Gharadig Basin > Abu Gharadig Field (0.99)
- (10 more...)
Experimental Investigation of Two-Phase Flow Properties of Heterogeneous Rocks Based on X-Ray Microfocus Radiography
Aรฉrens, P. (The University of Texas at Austin (currently with Halliburton)) | Espinoza, D. N. (The University of Texas at Austin) | Torres-Verdรญn, C. (The University of Texas at Austin (Corresponding author))
Summary An uncommon facet of formation evaluation is the assessment of flow-related in-situ properties of rocks. Most of the models used to describe two-phase flow properties of porous rocks assume homogeneous and/or isotropic media, which is hardly the case with actual reservoir rocks, regardless of scale; carbonates and grain-laminated sandstones are but two common examples of this situation. The degree of spatial complexity of rocks and its effect on the mobility of hydrocarbons are of paramount importance for the description of multiphase fluid flow in most contemporary reservoirs. There is thus a need for experimental and numerical methods that integrate all salient details about fluid-fluid and rock-fluid interactions. Such hybrid, laboratory-simulation projects are necessary to develop realistic models of fractional flow in complex rocks, i.e., saturation-dependent capillary pressure and relative permeability. Furthermore, these two crucial properties are usually measured independently. Capillary pressure is typically assessed using static measurements and unrealistic pressure conditions, whereas relative permeability is evaluated dynamically. Consequently, the disparity between the nature of the two experimental procedures often results in a potentially significant loss of information. We document a new high-resolution visualization technique that provides experimental insight to quantify fluid saturation patterns in heterogeneous rocks which allow for the simultaneous and dynamic evaluation of two-phase flow properties. The experimental apparatus consists of an X-ray microfocus scanner and an automated syringe pump. Rather than using traditional cylindrical cores, thin rectangular rock samples are examined, their thickness being one order of magnitude smaller than the remaining two dimensions. During the experiment, the core is scanned quasicontinuously while the fluids are being injected, allowing for time-lapse visualization of the flood front. Numerical simulations are then conducted to match the experimental data and quantify effective saturation-dependent relative permeability and capillary pressure. The experimental results indicate that flow patterns and in-situ saturations are highly dependent on the nature of the heterogeneity and bedding-plane orientation during both imbibition and drainage cycles. In homogeneous rocks, fluid displacement approaches piston-like behavior. The assessment of capillary pressure and relative permeability is performed by examining the time-lapse water saturation profiles resulting from fluid displacement. In spatially complex rocks, high-resolution time-lapse images reveal preferential flow paths along high-permeability sections and a lowered sweep efficiency. Our experimental procedure emphasizes that capillary pressure and transmissibility differences play an important role in fluid-saturation distribution and sweep efficiency at late times. The method is fast and reliable to assess mixing laws for fluid-transport properties of rocks in spatially complex formations.
- Geology > Rock Type > Sedimentary Rock > Carbonate Rock (0.47)
- Geology > Rock Type > Sedimentary Rock > Clastic Rock > Sandstone (0.39)
Understanding the effects of permafrost degradation through a multi-physics approach
Vosoughi, Ehsan (Institut National de la Recherche Scientifique (INRS)) | Giroux, Bernard (Institut National de la Recherche Scientifique (INRS)) | Duchesne, Mathieu J. (Geological Survey of Canada) | Dupuis, J. Christian (Universit Laval)
Permafrost is a multiphase porous media that can host matter in all three states (solid, liquid, and gas). The equilibrium between the states of matter within the pore space is largely driven by salinity, pressure, and temperature. The complex interactions between the different thermodynamic processes can lead to a complex pore system that is altered at each subsequent thaw and freeze cycle. The dynamic changes imposed on this porous media alter the mechanical and electrical properties of the samples. These changes can thus be quantified and monitored using ultrasonic and electrical resistivity measurements. The experimental results presented in this work document the impacts of a thawing event on unconsolidated quartz sand samples that were partially saturated with a brine solution. The electrical resistivity and ultrasonic data were acquired simultaneously throughout the experiment and the spatiotemporal changes within the solid matrix were captured by time-lapse X-ray Computed Tomography. A total of 39 different samples were investigated. The two independent variables chosen for this study were the grain size and the salinity of the brines. The results show a clear transition in electrical and elastic properties as the material in the pore space transitions between two different states. Further results show that these transitions are the result of the alteration of the pore network itself. Also, the study of P-wave velocity, ice fraction, and X-ray computed tomography of two different types of ice that coexist within the pore network is documented. Given the distinct impact of two different types of ice on this cryogenic porous media, it is imperative to thoroughly comprehend the existence of different ice types before undertaking the electro-elastic investigation of permafrost.
- Research Report > New Finding (0.86)
- Research Report > Experimental Study (0.54)
- Geophysics > Seismic Surveying (1.00)
- Geophysics > Borehole Geophysics (1.00)
- Energy > Oil & Gas > Upstream (1.00)
- Materials > Chemicals > Commodity Chemicals > Petrochemicals (0.92)
- Reservoir Description and Dynamics > Reservoir Characterization > Seismic processing and interpretation (1.00)
- Reservoir Description and Dynamics > Formation Evaluation & Management > Open hole/cased hole log analysis (1.00)
- Data Science & Engineering Analytics > Information Management and Systems > Artificial intelligence (0.92)
- Information Technology > Artificial Intelligence (0.92)
- Information Technology > Sensing and Signal Processing > Image Processing (0.45)
Coherence-enhancing anisotropic diffusion filter for 3D high-resolution reconstruction of P-wave velocity and density using full-waveform inversion: Application to a North Sea ocean bottom cable data set
Mรฉtivier, Ludovic (Universitรฉ Grenoble Alpes, Universitรฉ Grenoble Alpes) | Brossier, Romain (Universitรฉ Grenoble Alpes) | Hoffmann, Alexandre (Universitรฉ Grenoble Alpes) | Mirebeau, Jean-Marie (Universitรฉ Paris-Saclay) | Provenzano, Giuseppe (Universitรฉ Grenoble Alpes) | Tarayoun, Alizia (Universitรฉ Grenoble Alpes) | Yong, Peng (Universitรฉ Grenoble Alpes)
ABSTRACT Regularization is a central topic in the study of the solutions of ill-posed inverse problems. High-resolution seismic imaging using full-waveform inversion (FWI) belongs to this category of problems. Regularization through anisotropic diffusion, a technique that emerged in the field of image processing, is an interesting alternative to conventional regularization strategies. Exploiting the structural information of a given image, it has the capability to smooth this image along its main structures. The main difficulty is how to design the anisotropic diffusion operator. The concept of coherence enhancing proposed in 2D is extended in 3D and applied so as to filter and enhance the structural coherence of the model updates within an FWI algorithm. The benefits of this strategy are investigated on a 2D synthetic experiment before considering the multiparameter inversion of a 3D field data set from the North Sea up to 10ย Hz. From this data, the vertical velocity and the density are simultaneously reconstructed. Compared with a conventional nonstationary Gaussian regularization strategy, the models obtained using the coherence-enhancing anisotropic diffusion strategy indicate an enhanced resolution, especially for the density model. The high-resolution reflectivity image computed from the impedance volume clearly illustrates the benefit this filtering approach can deliver in terms of structural interpretation.
- Europe > United Kingdom > North Sea (0.60)
- Europe > Norway > North Sea (0.60)
- Europe > North Sea (0.60)
- (2 more...)
- Geophysics > Seismic Surveying > Seismic Processing (1.00)
- Geophysics > Seismic Surveying > Seismic Modeling > Velocity Modeling > Seismic Inversion (1.00)
- Europe > Norway > North Sea > Central North Sea > Central Graben > PL 033 > Block 2/11 > Hod Field > Tor Formation (0.94)
- Europe > Norway > North Sea > Central North Sea > Central Graben > PL 033 > Block 2/11 > Hod Field > Hod Formation (0.94)
- Europe > Norway > North Sea > Central North Sea > Central Graben > PL 033 > Block 2/11 > Hod Field > Ekofisk Formation (0.94)
- (8 more...)
- Information Technology > Artificial Intelligence (1.00)
- Information Technology > Sensing and Signal Processing > Image Processing (0.54)
TransInver: 3D data-driven seismic inversion based on self-attention
Li, Kewen (China University of Petroleum (East China) Qingdao) | Dou, Yimin (China University of Petroleum (East China) Qingdao) | Xiao, Yuan (China University of Petroleum (East China) Qingdao) | Jing, Ruilin (Shengli Oilfield Company) | Zhu, Jianbing (Shengli Oilfield Company) | Ma, Chengjie (Shengli Oilfield Company)
ABSTRACT Recently, convolutional neural network (CNN)-based deep learning (DL) for impedance inversion has been extended to multiple dimensions. Training multidimensional DL inversion requires extracting supervised information from sparse 1D well-log labels. Fully convolutional networks rely on their parameter sharing mechanism and receptive fields to achieve this, but their perceptual range is limited, making it difficult to capture long-term correlations in seismic data. The transformer is a type of network that is entirely based on self-attention, and it has demonstrated remarkable performance across various tasks and domains. However, its suitability for 3D seismic inversion is restricted by its high computational workload, fixed input size requirement, and inadequate handling of low-level details. The primary goal was to reengineer the self-attention mechanism to optimize its applicability for seismic impedance inversion tasks. The high-dimensional self-attention is decoupled into dual low-dimensional attention paths to reduce the computation of dense connections and matrix dot products. Shared parameters are used instead of full connection, allowing for flexible changes in input sizes by the network. In addition, its local modeling capabilities are enhanced by integrating it with the residual structure of the CNN. We name the resulting structure Self-Attention ResBlock, which is used as the basic unit for constructing TransInver. Comparative experiments indicate that TransInver performs significantly better than 3D methods such as UNet, TransUNet, HRNet, and 1D inversion methods. TransInver produces reliable inversion results using only nine well logs for SEAM Phase I and three well logs for the field data set of the Netherlands F3. This backbone network can deliver excellent inversion performance without depending on any auxiliary means such as low-frequency constraints or semisupervised frameworks.
- Geophysics > Seismic Surveying > Seismic Processing (1.00)
- Geophysics > Seismic Surveying > Seismic Modeling > Velocity Modeling > Seismic Inversion (1.00)
- Information Technology > Sensing and Signal Processing > Image Processing (1.00)
- Information Technology > Artificial Intelligence > Natural Language (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks > Deep Learning (1.00)