FaciesViT: Vision transformer for an improved core lithofacies prediction

Koeshidayatullah, Ardiansyah; Al-Azani, Sadam; Baraboshkin, E.Y.; Alfarraj, Motaz

doi:10.3389/feart.2022.992442

Cited by 17 publications

(7 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Despite these advances, it is worth noting that developing and training machine learning algorithms takes time and requires expertise outside of geosciences to conduct the necessary data pre-conditioning to run such machine learning models 16 , 23 – 25 . Furthermore, time-consuming processes such as data preparation and processing, selecting appropriate parameters, and fine-tuning the model are frequently required to test and compare different classification algorithms.…”

Section: Introductionmentioning

confidence: 99%

“…To overcome this issue, several works have proposed different workflows to automate model generation, tuning and evaluation processes, or to create an automated machine learning (AutoML) approach 26 , 27 . In such a case, AutoML focuses on hyperparameter optimization and model optimization by using Bayesian optimization, genetic algorithms, or reinforcement learning 25 .…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Hierarchical automated machine learning (AutoML) for advanced unconventional reservoir characterization

Mubarak,

Koeshidayatullah

2023

Sci Rep

Self Cite

View full text Add to dashboard Cite

Recent advances in machine learning (ML) have transformed the landscape of energy exploration, including hydrocarbon, CO2 storage, and hydrogen. However, building competent ML models for reservoir characterization necessitates specific in-depth knowledge in order to fine-tune the models and achieve the best predictions, limiting the accessibility of machine learning in geosciences. To mitigate this issue, we implemented the recently emerged automated machine learning (AutoML) approach to perform an algorithm search for conducting an unconventional reservoir characterization with a more optimized and accessible workflow than traditional ML approaches. In this study, over 1000 wells from Alberta’s Athabasca Oil Sands were analyzed to predict various key reservoir properties such as lithofacies, porosity, volume of shale, and bitumen mass percentage. Our proposed workflow consists of two stages of AutoML predictions, including (1) the first stage focuses on predicting the volume of shale and porosity by using conventional well log data, and (2) the second stage combines the predicted outputs with well log data to predict the lithofacies and bitumen percentage. The findings show that out of the ten different models tested for predicting the porosity (78% in accuracy), the volume of shale (80.5%), bitumen percentage (67.3%), and lithofacies classification (98%), distributed random forest, and gradient boosting machine emerged as the best models. When compared to the manually fine-tuned conventional machine learning algorithms, the AutoML-based algorithms provide a notable improvement on reservoir property predictions, with higher weighted average f1-scores of up to 15–20% in the classification problem and 5–10% in the adjusted-R2 score for the regression problems in the blind test dataset, and it is achieved only after ~ 400 s of training and testing processes. In addition, from the feature ranking extraction technique, there is a good agreement with domain experts regarding the most significant input parameters in each prediction. Therefore, it is evidence that the AutoML workflow has proven powerful in performing advanced petrophysical analysis and reservoir characterization with minimal time and human intervention, allowing more accessibility to domain experts while maintaining the model’s explainability. Integration of AutoML and subject matter experts could advance artificial intelligence technology implementation in optimizing data-driven energy geosciences.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Hierarchical automated machine learning (AutoML) for advanced unconventional reservoir characterization

Mubarak,

Koeshidayatullah

2023

Sci Rep

Self Cite

View full text Add to dashboard Cite

show abstract

“…Its ability to train DNN effectively, while mitigating common ML challenges, positions ResNet as a superior choice for capturing intricate relationships between CWT seismic components and identified facies . ResNet obtained impressive importance in advancing the field of deep learning (DL) by incorporating image identification, natural language processing, and machine translation , while alleviating issues like resilient to noise and overfitting, vanishing gradients, and being more efficient and less computationally expensive to train. − …”

Section: Introductionmentioning

confidence: 99%

ResNet and CWT Fusion: A New Paradigm for Optimized Heterogeneous Thin Reservoir Evaluation

Akram,

Akhter,

et al. 2024

ACS Omega

View full text Add to dashboard Cite

The endeavor to explore and characterize oil and gas reservoirs presents significant challenges due to the inherent heterogeneities that are further compounded by the existence of thin sand layers encapsulated in shale strata. This complexity is intensified by limited and low-resolution seismic data, missing critical well-log information, and inaccessible angle stack data. Conventional reservoir classification approaches have struggled to address these issues, primarily due to their limitations in handling missing data effectively and, hence, precise estimations. This study focuses on the characterization of thin, heterogeneous potential sands of the B-interval within the Lower Goru Formation, a proven gas reservoir in the Badin area. The reservoir sands with varying thicknesses are assessed in detail for their optimized description and field productions by handling challenges, including low seismic resolutions, heterogeneities, and missing data sets. An innovative solution is developed based on the integration of continuous wavelet transform (CWT) and machine learning (ML) techniques for the approximation of missing data sets, i.e., S-wave (DTS), along with enhanced elastic and petrophysical properties. The improved properties are augmented by the high resolution attained by CWT and captured variability more profoundly through the implication of residual neural networks (ResNet). The limitations of conventional approaches are harnessed by ML solutions that operate with limited input data and deliver significantly improved results in characterizing enigmatic thin sand reservoirs. The high-frequency petroelastic properties reliably determined the thin heterogeneous potential sand bodies and illuminated a channelized play fairway that can be tested for additional wells with low-risk involvement.

show abstract

“…Liang et al (2021) first used a ViT network structure that evolved from transformers to classify seven different types of ores. Koeshidayatullah et al (2022) proposed a novel FaciesViT model based on the transformer framework for automatic core facies classification, which is much better than CNN and hybrid CNN-ViT models, and does not require preprocessing and feature extraction. In addition to rock images of natural scenes, many scholars also use microscopic rock images and spectral images for rock classification.…”

Section: Introductionmentioning

confidence: 99%

Rock image classification using deep residual neural network with transfer learning

Chen

Su²,

Chen³

et al. 2023

Front. Earth Sci.

View full text Add to dashboard Cite

Rock image classification is a significant part of geological research. Compared with traditional image classification methods, rock image classification methods based on deep learning models have the great advantage in terms of automatic image features extraction. However, the rock classification accuracies of existing deep learning models are unsatisfied due to the weak feature extraction ability of the network model. In this study, a deep residual neural network (ResNet) model with the transfer learning method is proposed to establish the corresponding rock automatic classification model for seven kinds of rock images. ResNet34 introduces the residual structure to make it have an excellent effect in the field of image classification, which extracts high-quality rock image features and avoids information loss. The transfer learning method abstracts the deep features from the shallow features, and better express the rock texture features for classification in the case of fewer rock images. To improve the generalization of the model, a total of 3,82,536 rock images were generated for training via image slicing and data augmentation. The network parameters trained on the Texture Library dataset which contains 47 types of texture images and reflect the characteristics of rocks are used for transfer learning. This pre-trained weight is loaded when training the ResNet34 model with the rock dataset. Then the model parameters are fine-tuned to transfer the model to the rock classification problem. The experimental results show that the accuracy of the model without transfer learning reached 88.1%, while the model using transfer learning achieved an accuracy of 99.1%. Aiming at geological engineering field investigation, this paper studies the embedded deployment application of the rock classification network. The proposed rock classification network model is transplanted to an embedded platform. By designing a rock classification system, the off-line rock classification is realized, which provides a new solution for the rock classification problem in the geological survey. The deep residual neural network and transfer learning method used in this paper can automatically classify rock features without manually extracting. These methods reduce the influence of subjective factors and make the rock classification process more automatic and intelligent.

show abstract

FaciesViT: Vision transformer for an improved core lithofacies prediction

Cited by 17 publications

References 56 publications

Hierarchical automated machine learning (AutoML) for advanced unconventional reservoir characterization

Hierarchical automated machine learning (AutoML) for advanced unconventional reservoir characterization

ResNet and CWT Fusion: A New Paradigm for Optimized Heterogeneous Thin Reservoir Evaluation

Rock image classification using deep residual neural network with transfer learning

Contact Info

Product

Resources

About