Abstract:One of the key and difficult points in seismic data processing is seismic data denoising. Under the influence of the acquisition environment, the collected seismic data usually have low signal-to-noise ratio (SNR) and low resolution. In desert region of western China, different from other regions, the random noise there has complex characteristics of non-Gaussian, non-stationary and non-linearity. Its main frequency is quite low and its spectral overlap with those of effective signals. Moreover, large amount of data and intelligent requirements make traditional denoising methods encounter difficulties. In addition, the existing deep learning denoising methods also reveal two problems: first, they can only effectively suppress simple seismic noise such as Gaussian noise. But for the case of desert noise, they usually mistakenly judge noise as effective signals and destroy signal characteristic structure. Second, what is more important, in supervised-based methods paired noisy and pure data are usually used as training set. However, there is no noise-free data in actual desert seismic data, which severely limits the processing performance of supervised-based networks. Thus, a new parameter-shared variational auto-encoding adversarial network (PS-VAAN) is proposed for desert seismic data denoising in this paper. This new method includes two encoders, two generators and two discriminators. It can realize domain conversion from noisy data domain to pure data domain. We construct relatively complete unpaired pure and noisy data training sets for training, and design reconstruction loss function and adversarial loss function to optimize network parameters. Moreover, cycle-consistency loss is introduced to make sure two domains mapping into the same space. Compared with the state-of-art denoising methods on synthetic and actual seismic records, the proposed method not only has superior denoising ability, but also has strong feature extraction and fitting ability to recover effective signals with almost no energy loss.

Intelligent Random Noise Modeling by the Improved Variational Autoencoding Method and Its Application to Data Augmentation

Parameter-shared variational auto-encoding adversarial network for desert seismic data denoising in Northwest China

Random Seismic Noise Attenuation Based on Data Augmentation and CNN

Noise Modeling and Data Augmentation Using Conditional Adversarial Autoencoder

Seismic Random Noise Attenuation Based on Non-IID Pixel-Wise Gaussian Noise Modeling

Seismic random noise suppression using deep convolutional autoencoder neural network

Modeling Land Seismic Exploration Random Noise in a Weakly Heterogeneous Medium and the Application to the Training Set

Data Augmentation and Its Application in Distributed Acoustic Sensing Data Denoising

Seismic Shot Gather Denoising by Using a Supervised-Deep-Learning Method with Weak Dependence on Real Noise Data: A Solution to the Lack of Real Noise Data

Unsupervised Deep Learning for Random Noise Attenuation of Seismic Data

Unsupervised Erratic Seismic Noise Attenuation with Robust Deep Convolutional Autoencoders

SeNM-VAE: Semi-Supervised Noise Modeling with Hierarchical Variational Autoencoder

Hologram Noise Model for Data Augmentation and Deep Learning

Time Series Data Augmentation Algorithm Combining Deep Metric Learning and Variational Encoder

Random and Coherent Noise Suppression in DAS-VSP Data by Using a Supervised Deep Learning Method

Unsupervised Seismic Random Noise Suppression Based on Local Similarity and Replacement Strategy.

Seismic Labeled Data Expansion Using Variational Autoencoders

Denoising of seismic data in desert environment based on a variational mode decomposition and a convolutional neural network

Deep Denoising Autoencoder for Seismic Random Noise Attenuation

Robust VAEs via Generating Process of Noise Augmented Data

Unpaired training: Optimize the seismic data denoising model without paired training data