arXiv subjects

Kewen Li

Publications and source records attributed to Kewen Li.

7 recordsLinked to original sources

arXiv2023

FaultSSL: Seismic Fault Detection via Semi-supervised learning

The prevailing methodology in data-driven fault detection leverages synthetic data for training neural networks. However, it grapples with challenges when it comes to generalization in surveys exhibiting complex structures. To enhance the generalization of models trained on limited synthetic datasets to a broader range of real-world data, we introduce FaultSSL, a semi-supervised fault detection framework. This method is based on the classical mean teacher structure, with its supervised part employs synthetic data and a few 2D labels. The unsupervised component relying on two meticulously devised proxy tasks, allowing it to incorporate vast unlabeled field data into the training process. The two proxy tasks are PaNning Consistency (PNC) and PaTching Consistency (PTC). PNC emphasizes the feature consistency of the overlapping regions between two adjacent views in predicting the model. This allows for the extension of 2D slice labels to the global seismic volume. PTC emphasizes the spatially consistent nature of faults. It ensures that the predictions for the seismic, whether made on the entire volume or on individual patches, exhibit coherence without any noticeable artifacts at the patch boundaries. While the two proxy tasks serve different objectives, they uniformly contribute to the enhancement of performance. Experiments showcase the exceptional performance of FaultSSL. In surveys where other mainstream methods fail to deliver, we present reliable, continuous, and clear detection results. FaultSSL breaks the shackles of synthetic data, unveiling a promising route for incorporating copious amounts of field data into training and fostering model generalization across a broader spectrum of surveys.

physics.geo-ph↗

arXiv2023

ContrasInver: Ultra-Sparse Label Semi-supervised Regression for Multi-dimensional Seismic Inversion

The automated interpretation and inversion of seismic data have advanced significantly with the development of Deep Learning (DL) methods. However, these methods often require numerous costly well logs, limiting their application only to mature or synthetic data. This paper presents ContrasInver, a method that achieves seismic inversion using as few as two or three well logs, significantly reducing current requirements. In ContrasInver, we propose three key innovations to address the challenges of applying semi-supervised learning to regression tasks with ultra-sparse labels. The Multi-dimensional Sample Generation (MSG) technique pioneers a paradigm for sample generation in multi-dimensional inversion. It produces a large number of diverse samples from a single well, while establishing lateral continuity in seismic data. MSG yields substantial improvements over current techniques, even without the use of semi-supervised learning. The Region-Growing Training (RGT) strategy leverages the inherent continuity of seismic data, effectively propagating accuracy from closer to more distant regions based on the proximity of well logs. The Impedance Vectorization Projection (IVP) vectorizes impedance values and performs semi-supervised learning in a compressed space. We demonstrated that the Jacobian matrix derived from this space can filter out some outlier components in pseudo-label vectors, thereby solving the value confusion issue in semi-supervised regression learning. In the experiments, ContrasInver achieved state-of-the-art performance in the synthetic data SEAM I. In the field data with two or three well logs, only the methods based on the components proposed in this paper were able to achieve reasonable results. It's the first data-driven approach yielding reliable results on the Netherlands F3 and Delft, using only three and two well logs respectively.

physics.geo-ph↗

arXiv2022

CONSS: Contrastive Learning Approach for Semi-Supervised Seismic Facies Classification

Recently, seismic facies classification based on convolutional neural networks (CNN) has garnered significant research interest. However, existing CNN-based supervised learning approaches necessitate massive labeled data. Labeling is laborious and time-consuming, particularly for 3D seismic data volumes. To overcome this challenge, we propose a semi-supervised method based on pixel-level contrastive learning, termed CONSS, which can efficiently identify seismic facies using only 1% of the original annotations. Furthermore, the absence of a unified data division and standardized metrics hinders the fair comparison of various facies classification approaches. To this end, we develop an objective benchmark for the evaluation of semi-supervised methods, including self-training, consistency regularization, and the proposed CONSS. Our benchmark is publicly available to enable researchers to objectively compare different approaches. Experimental results demonstrate that our approach achieves state-of-the-art performance on the F3 survey.

cs.CV↗

arXiv2022

MDA GAN: Adversarial-Learning-based 3-D Seismic Data Interpolation and Reconstruction for Complex Missing

The interpolation and reconstruction of missing traces is a crucial step in seismic data processing, moreover it is also a highly ill-posed problem, especially for complex cases such as high-ratio random discrete missing, continuous missing and missing in fault-rich or salt body surveys. These complex cases are rarely mentioned in current works. To cope with complex missing cases, we propose Multi-Dimensional Adversarial GAN (MDA GAN), a novel 3-D GAN framework. It keeps anisotropy and spatial continuity of the data after 3D complex missing reconstruction using three discriminators. The feature stitching module is designed and embedded in the generator to retain more information of the input data. The Tanh cross entropy (TCE) loss is derived, which provides the generator with the optimal reconstruction gradient to make the generated data smoother and continuous. We experimentally verified the effectiveness of the individual components of the study and then tested the method on multiple publicly available data. The method achieves reasonable reconstructions for up to 95% of random discrete missing and 100 traces of continuous missing. In fault and salt body enriched surveys, MDA GAN still yields promising results for complex cases. Experimentally it has been demonstrated that our method achieves better performance than other methods in both simple and complex cases.https://github.com/douyimin/MDA_GAN

physics.geo-ph↗

arXiv2021

MD Loss: Efficient Training of 3D Seismic Fault Segmentation Network under Sparse Labels by Weakening Anomaly Annotation

Data-driven fault detection has been regarded as a 3D image segmentation task. The models trained from synthetic data are difficult to generalize in some surveys. Recently, training 3D fault segmentation using sparse manual 2D slices is thought to yield promising results, but manual labeling has many false negative labels (abnormal annotations), which is detrimental to training and consequently to detection performance. Motivated to train 3D fault segmentation networks under sparse 2D labels while suppressing false negative labels, we analyze the training process gradient and propose the Mask Dice (MD) loss. Moreover, the fault is an edge feature, and current encoder-decoder architectures widely used for fault detection (e.g., U-shape network) are not conducive to edge representation. Consequently, Fault-Net is proposed, which is designed for the characteristics of faults, employs high-resolution propagation features, and embeds MultiScale Compression Fusion block to fuse multi-scale information, which allows the edge information to be fully preserved during propagation and fusion, thus enabling advanced performance via few computational resources. Experimental demonstrates that MD loss supports the inclusion of human experience in training and suppresses false negative labels therein, enabling baseline models to improve performance and generalize to more surveys. Fault-Net is capable to provide a more stable and reliable interpretation of faults, it uses extremely low computational resources and inference is significantly faster than other models. Our method indicates optimal performance in comparison with several mainstream methods.

cs.CV↗

arXiv2021

Attention-Based 3D Seismic Fault Segmentation Training by a Few 2D Slice Labels

Detection faults in seismic data is a crucial step for seismic structural interpretation, reservoir characterization and well placement. Some recent works regard it as an image segmentation task. The task of image segmentation requires huge labels, especially 3D seismic data, which has a complex structure and lots of noise. Therefore, its annotation requires expert experience and a huge workload. In this study, we present lambda-BCE and lambda-smooth L1loss to effectively train 3D-CNN by some slices from 3D seismic data, so that the model can learn the segmentation of 3D seismic data from a few 2D slices. In order to fully extract information from limited data and suppress seismic noise, we propose an attention module that can be used for active supervision training and embedded in the network. The attention heatmap label is generated by the original label, and letting it supervise the attention module using the lambda-smooth L1loss. The experiment demonstrates the effectiveness of our loss function, the method can extract 3D seismic features from a few 2D slice labels. And it also shows the advanced performance of the attention module, which can significantly suppress the noise in the seismic data while increasing the model's sensitivity to the foreground. Finally, on the public test set, we only use the 2D slice labels training that accounts for 3.3% of the 3D volume label, and achieve similar performance to the 3D volume label training.

cs.CV↗

arXiv2020

Prediction and analysis of Coronavirus Disease 2019

In December 2019, a novel coronavirus was found in a seafood wholesale market in Wuhan, China. WHO officially named this coronavirus as COVID-19. Since the first patient was hospitalized on December 12, 2019, China has reported a total of 78,824 confirmed CONID-19 cases and 2,788 deaths as of February 28, 2020. Wuhan's cumulative confirmed cases and deaths accounted for 61.1% and 76.5% of the whole China mainland , making it the priority center for epidemic prevention and control. Meanwhile, 51 countries and regions outside China have reported 4,879 confirmed cases and 79 deaths as of February 28, 2020. COVID-19 epidemic does great harm to people's daily life and country's economic development. This paper adopts three kinds of mathematical models, i.e., Logistic model, Bertalanffy model and Gompertz model. The epidemic trends of SARS were first fitted and analyzed in order to prove the validity of the existing mathematical models. The results were then used to fit and analyze the situation of COVID-19. The prediction results of three different mathematical models are different for different parameters and in different regions. In general, the fitting effect of Logistic model may be the best among the three models studied in this paper, while the fitting effect of Gompertz model may be better than Bertalanffy model. According to the current trend, based on the three models, the total number of people expected to be infected is 49852-57447 in Wuhan,12972-13405 in non-Hubei areas and 80261-85140 in China respectively. The total death toll is 2502-5108 in Wuhan, 107-125 in Non-Hubei areas and 3150-6286 in China respetively. COVID-19 will be over p robably in late-April, 2020 in Wuhan and before late-March, 2020 in other areas respectively.

q-bio.PE↗