Seismic acoustic impedance (AI) inversion is essential for reservoir prediction and characterization. In recent years, deep learning has shown immense potential as a data-driven approach in seismic data processing, inversion, and interpretation. As a data-driven method, deep learning-based seismic inversion results better when sufficient labeled data are provided. Overfitting and poor generalization often occur when labels are insufficient. Due to the lack of labeled data in seismic inversion problems, the difficulty of inversion increases, leading to unstable and poor generalization of prediction results. To partially address this issue, we propose a constrained seismic inversion strategy. Since seismic records are time series, we exploit the convolutional neural network (CNN) and bidirectional LSTM (Bi-LSTM) network structures that are more applicable to time series. We combine the physical model and the initial model as constraints to improve the network stability and generalization ability, and impose sparse constraints on the reflection coefficient to further improve the prediction accuracy. The network structure transformation improves the efficiency and stability of the training process. Through numerical experiments and real data tests, it is proved that the proposed method improves the vertical resolution and geological reliability, providing a more stable and efficient method for seismic inversion under conditions of limited labeled data. The overall performance improved by 2% through comparative analysis.

Affected by near-surface absorption, seismic wave energy attenuation and phase distortion greatly reduce the resolution and signal-to-noise ratio (SNR) of seismic data, causing changes in seismic attributes much greater than other factors. Inverse Q filtering is a common method to compensate for these undesirable effects. To overcome the drawbacks of the traditional inverse Q filtering, such as the difficulty of parameter selection and the instability of wave amplitude compensation, we propose a new unsupervised inverse Q filtering method in a deep learning (DL) framework, using a forward attenuation operator based on the seismic wave attenuation theory to drive the network. The filtering strategy does not require actual training labels and avoids the numerical instability of the amplitude compensation. First, we design a hybrid convolutional neural network bidirectional LSTM (CNN-BiLSTM)-attention model for multivariate time series prediction and then take the data to be compensated as input for the DL network and the compensated data as output. The output is then attenuated using a forward attenuation operator constructed from the near-surface Q model. After that, the error between the attenuated data and the original input data is transmitted back to the DL network to modify the network output, and the error is minimized by optimizing the network parameters to generate the final compensation result. In the entire prediction process, there is no need to produce unattenuated data labels, which achieves the effect of unsupervised learning. The results with synthetic and field data demonstrate that the unsupervised method can effectively and stably compensate for seismic signals. Compared to the classical inverse Q filtering, the proposed method improves the resolution and SNR of seismic records.

Deep learning has made substantial progress in remote sensing image segmentation tasks. It usually requires a large number of high-quality annotation maps (i.e., clean labels), which are labor-intensive. In this paper, we propose to use abundant electronic maps (i.e., noisy labels) to supplement a small number of clean labels to solve the problem of massive label production. In addition, we propose a multi-stage noise supervised framework (NSDI) to prevent noise from deteriorating the performance of deep model. NSDI consists ofclean training,weight initialization, andhybrid trainingstages. In theclean trainingstage, we train a segmentation model using a small amount of clean labels and remote sensing images to compute the confusion probability matrix. In theweight initializationstage, we use the confusion probability matrix as well as the prediction probability to calculate the label reliability of the electronic map. In thehybrid trainingstage, an adaptive weighted loss function based on cross-entropy is used to dynamically update the label reliability. Then we train model further using electronic maps with the support of the adaptive weighted loss function and label reliability. Experiments were undertaken on 2,656 images of 512 × 512 pixels. Ablation studies show that NSDI improves the model robustness as well as the segmentation quality.

Brittleness is an important parameter characterizing the fracturing properties of shale reservoir, which can be predicted by the pre-stack seismic inversion. In order to overcome the low efficiency and ill-posed problems of the traditional pre-stack brittleness inversion, we propose a new unsupervised deep learning (DL) inversion method for seismic brittleness parameters based on the physical equation. This method integrates DL framework and the physical equation, and provides a DL inversion strategy without actual labels. We first input the original seismic data into the Fastformer network, and use the low-frequency model as the physical constraint to predict the brittle parameters. Then, the prediction results of brittleness parameters are sent to the forward modeling module (a linear approximation equation) to calculate the synthetic seismic data. Next, the error between the calculated seismic data and original seismic data is used to update the network prediction results. The network parameters are iteratively optimized to minimize the error, and the brittle prediction parameters are finally output. In the whole training process, it is not necessary to use the real brittle parameters as the labels. Through this method, the effect of approximate unsupervised learning is obtained. Finally, we apply the proposed method to the synthetic data and field data, and compared with the results inverted by the traditional L1 method. The experimental results show that the proposed method has higher inversion accuracy and efficiency than the traditional L1 method, which has a great potential in the practical application.

Due to low cost and nondestructive characteristics, the passive surface-wave imaging has shown great potentials in urban near-surface exploration. However, the imaging methods are facing with many challenges in practical applications, such as uneven noise source distribution and complex site environment, which will seriously affect the dispersion imaging quality of surface waves, and result in the failure of retrieving accurate dispersion curve and inaccurate inversion. Therefore, data screening is required to improve the accuracy of passive surface-wave dispersion imaging. This process is usually completed manually, which is time-consuming for processing the large data sets. To solve this problem, we propose an automatic data screening method for the near-surface passive surface-wave imaging. Based on the distribution characteristics of the surface-wave energy of noise data in the F-K domain, this proposed method uses the least-squares technique to fit the quadratic distribution of energy, and sorts the noise time segments according to the defined correlation coefficients and bandwidth coefficients. Thus, we can automatically detect the noise time segments with high signal-to-noise ratio (SNR) without manual intervention. In order to verify the effectiveness of this proposed method, both synthetic data and field data are used in this study. The results show that this automatic data screening method significantly improves the accuracy of the passive surface-wave dispersion imaging, effectively expands the surface-wave energy band, and realizes rapid and automatic data screening.

Surface-wave analysis methods have been widely applied to construct near-surface shear-wave velocity structures. Whether it is an active source or passive source, the near-surface shear-wave velocity structure is obtained by inverting the surface-wave dispersion curve. In order to solve the problems of low inversion efficiency and poor inversion results in traditional surface-wave exploration, we have studied the surface-wave inversion methods based on deep learning technology. In this study, we propose a long short-term memory (LSTM) surface-wave inversion method based on the first height last velocity (FHLV) loss function. The core of our proposed method is the FHLV loss function consisting of two parts: a speed loss and a thickness loss, which improves the overall prediction accuracy through optimizing the learning process of the thickness parameter by the network. To verify the accuracy of the proposed LSTM surface-wave inversion method based on the FHLV loss function, experiments are conducted on both synthetic and real datasets. The results show that our proposed method can efficiently and accurately invert the near-surface shear-wave velocity structure.