Abstract:To limit the time, cost, and environmental impact associated with the acquisition of seismic data, in recent decades considerable effort has been put into so-called simultaneous shooting acquisitions, where seismic sources are fired at short time intervals between each other. As a consequence, waves originating from consecutive shots are entangled within the seismic recordings, yielding so-called blended data. For processing and imaging purposes, the data generated by each individual shot must be retrieved. This process, called deblending, is achieved by solving an inverse problem which is heavily underdetermined. Conventional approaches rely on the action of the adjoint of the blending operator that renders the blending noise into burst-like noise, whilst preserving the signal of interest. Compressed sensing type regularization is then applied, where sparsity in some domain is assumed for the signal of interest. The domain of choice generally depends on the acquisition geometry and the separability between signal and noise within the chosen domain. In this work, we introduce a new concept that consists of embedding a self-supervised denoising network into the Plug-and-Play (PnP) framework. A novel network is introduced whose design extends an existing blind-spot network architecture for partially coherent noise (i.e., correlated in time). The network is then trained directly on the noisy input data at each step of the PnP algorithm. By leveraging both the underlying physics of the problem and the great denoising capabilities of our blind-spot network, the proposed algorithm is shown to outperform industry-standard methods whilst being comparable in terms of computational cost. Moreover, being independent of the acquisition geometry, our method can be easily applied to both marine and land data without any significant modification.

The potential of self-supervised networks for random noise suppression in seismic data

Self-Noise Suppression for AUV without Clean Data: a Noise2Noise Approach

Coherent noise suppression via a self-supervised blind-trace deep learning scheme

Seismic Data Random Noise Attenuation Using Visible Blind Spot Self-Supervised Learning

Unsupervised Seismic Data Random Noise Attenuation Method Based on Improved Blind Spot Strategy

Trace-wise coherent noise suppression via a self-supervised blind-trace deep learning scheme

A Denoising Method of Seismic Data Based on Self-Supervised Learning

Self-Supervised Seismic Random Noise Attenuation With Spatial Attention From a Single Section

Seismic Random Noise Attenuation Via Self-Supervised Transfer Learning

A U-Net Based Deep Learning Approach for Seismic Random Noise Suppression

Self-Supervised Denoising for Distributed Acoustic Sensing Vertical Seismic Profile Data via Improved Blind Spot Network

Random Noise Suppression of Seismic Data Through Multi-Scale Residual Dense Network

Unsupervised Deep Learning for Random Noise Attenuation of Seismic Data

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

An effective self-supervised learning method for various seismic noise attenuation

Integrating self-supervised denoising in inversion-based seismic deblending

A Natural Images Pre-Trained Deep Learning Method for Seismic Random Noise Attenuation

A self-supervised scheme for ground roll suppression

Random Noise Attenuation by Self-supervised Learning from Single Seismic Data

A hybrid approach to seismic deblending: when physics meets self-supervision

Noise Suppression for CRP Gathers Based on Self2Self with Dropout