CDDIP: Constrained Diffusion-Driven Deep Image Prior for Seismic Image Reconstruction

TL;DR

This paper introduces a novel seismic image reconstruction method combining a pre-trained diffusion model with Deep Image Prior, achieving robust, high-quality results with reduced inference time even on data outside the training domain.

Contribution

It presents a new approach that integrates diffusion models and Deep Image Prior for seismic data reconstruction, improving robustness and efficiency over existing methods.

Findings

Effective reconstruction of missing seismic traces in synthetic and field data.

Reduces sampling timesteps during inference by up to 4x.

Handles high geological variability and out-of-domain data.

Abstract

Seismic data frequently exhibits missing traces, substantially affecting subsequent seismic processing and interpretation. Deep learning-based approaches have demonstrated significant advancements in reconstructing irregularly missing seismic data through supervised and unsupervised methods. Nonetheless, substantial challenges remain, such as generalization capacity and computation time cost during the inference. Our work introduces a reconstruction method that uses a pre-trained generative diffusion model for image synthesis and incorporates Deep Image Prior to enforce data consistency when reconstructing missing traces in seismic data. The proposed method has demonstrated strong robustness and high reconstruction capability of post-stack and pre-stack data with different levels of structural complexity, even in field and synthetic scenarios where test data were outside the training domain. This indicates that our method can handle the high geological variability of different exploration targets. Additionally, compared to other state-of-the-art seismic reconstruction methods using diffusion models. During inference, our approach reduces the number of sampling timesteps by up to 4x.