Self-Supervised Annotation of Seismic Images using Latent Space\n Factorization

2020/09/09 by Oluwaseun Joseph Aribido, Aribido, Oluwaseun Joseph, Ghassan AlRegib +3 · 1 citation
Earth and Planetary Sciences · Engineering · #Drilling and Well Engineering #FOS: Computer and information sciences #FOS: Electrical engineering #Hydraulic Fracturing and Reservoir Analysis #Image and Video Processing (eess.IV) #Machine Learning (cs.LG) #Seismic Imaging and Inversion Techniques #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.2009.04631

openalex publication_date 2020/09/09 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Annotating seismic data is expensive, laborious and subjective due to the\nnumber of years required for seismic interpreters to attain proficiency in\ninterpretation. In this paper, we develop a framework to automate annotating\npixels of a seismic image to delineate geological structural elements given\nimage-level labels assigned to each image. Our framework factorizes the latent\nspace of a deep encoder-decoder network by projecting the latent space to\nlearned sub-spaces. Using constraints in the pixel space, the seismic image is\nfurther factorized to reveal confidence values on pixels associated with the\ngeological element of interest. Details of the annotated image are provided for\nanalysis and qualitative comparison is made with similar frameworks.\n

Cited by

• A Large-scale Benchmark on Geological Fault Delineation Models: Domain Shift, Training Dynamics, Generalizability, Evaluation and Inferential Behavior

Related

• Self-Supervised Delineation of Geological Structures using Orthogonal Latent Space Projection
• Automatic classification of geologic units in seismic images using partially interpreted examples
• Unsupervised seismic facies classification using deep convolutional autoencoder
• CONSS: Contrastive Learning Approach for Semi-Supervised Seismic Facies Classification
• Encoding the Subsurface in 3D with Seismic