Revolutionizing seismic data compression: unlocking the power of stable diffusion neural networks

Ayrat Abdullin; Umair Bin Waheed; Naveed Iqbal

doi:10.1190/image2023-3911264.1

Revolutionizing seismic data compression: unlocking the power of stable diffusion neural networks

Ayrat Abdullin^*, Umair Bin Waheed, Naveed Iqbal

^*Corresponding author for this work

Research output: Contribution to journal › Conference article › peer-review

Abstract

Seismic data compression is a significant concern with the growing interest in distributed acoustic sensing. The large amounts of data generated by fiber optic sensors must be processed and analyzed in real-time, requiring the use of sophisticated compression techniques. However, achieving high compression ratios without sacrificing accuracy or speed remains a challenge. We propose a novel implementation of a stable diffusion (SD) deep neural network for in-field seismic data compression. We show that the proposed SD-based method achieves a compression ratio of 307 on a synthetic data set.

Original language	English
Pages (from-to)	474-478
Number of pages	5
Journal	SEG Technical Program Expanded Abstracts
Volume	2023-August
DOIs	https://doi.org/10.1190/image2023-3911264.1
State	Published - 14 Dec 2023
Event	3rd International Meeting for Applied Geoscience and Energy, IMAGE 2023 - Houston, United States Duration: 28 Aug 2023 → 1 Sep 2023

Bibliographical note

Publisher Copyright:
© 2023 Society of Exploration Geophysicists and the American Association of Petroleum Geologists.

ASJC Scopus subject areas

Geotechnical Engineering and Engineering Geology
Geophysics

Access to Document

10.1190/image2023-3911264.1

Cite this

@article{bdb1cc0721eb409a90afbd27e2f4041a,

title = "Revolutionizing seismic data compression: unlocking the power of stable diffusion neural networks",

abstract = "Seismic data compression is a significant concern with the growing interest in distributed acoustic sensing. The large amounts of data generated by fiber optic sensors must be processed and analyzed in real-time, requiring the use of sophisticated compression techniques. However, achieving high compression ratios without sacrificing accuracy or speed remains a challenge. We propose a novel implementation of a stable diffusion (SD) deep neural network for in-field seismic data compression. We show that the proposed SD-based method achieves a compression ratio of 307 on a synthetic data set.",

author = "Ayrat Abdullin and Waheed, {Umair Bin} and Naveed Iqbal",

note = "Publisher Copyright: {\textcopyright} 2023 Society of Exploration Geophysicists and the American Association of Petroleum Geologists.; 3rd International Meeting for Applied Geoscience and Energy, IMAGE 2023 ; Conference date: 28-08-2023 Through 01-09-2023",

year = "2023",

month = dec,

day = "14",

doi = "10.1190/image2023-3911264.1",

language = "English",

volume = "2023-August",

pages = "474--478",

journal = "SEG Technical Program Expanded Abstracts",

issn = "1052-3812",

publisher = "Society of Exploration Geophysicists",

}

TY - JOUR

T1 - Revolutionizing seismic data compression

T2 - 3rd International Meeting for Applied Geoscience and Energy, IMAGE 2023

AU - Abdullin, Ayrat

AU - Waheed, Umair Bin

AU - Iqbal, Naveed

PY - 2023/12/14

Y1 - 2023/12/14

N2 - Seismic data compression is a significant concern with the growing interest in distributed acoustic sensing. The large amounts of data generated by fiber optic sensors must be processed and analyzed in real-time, requiring the use of sophisticated compression techniques. However, achieving high compression ratios without sacrificing accuracy or speed remains a challenge. We propose a novel implementation of a stable diffusion (SD) deep neural network for in-field seismic data compression. We show that the proposed SD-based method achieves a compression ratio of 307 on a synthetic data set.

AB - Seismic data compression is a significant concern with the growing interest in distributed acoustic sensing. The large amounts of data generated by fiber optic sensors must be processed and analyzed in real-time, requiring the use of sophisticated compression techniques. However, achieving high compression ratios without sacrificing accuracy or speed remains a challenge. We propose a novel implementation of a stable diffusion (SD) deep neural network for in-field seismic data compression. We show that the proposed SD-based method achieves a compression ratio of 307 on a synthetic data set.

UR - http://www.scopus.com/inward/record.url?scp=85180533997&partnerID=8YFLogxK

U2 - 10.1190/image2023-3911264.1

DO - 10.1190/image2023-3911264.1

M3 - Conference article

AN - SCOPUS:85180533997

SN - 1052-3812

VL - 2023-August

SP - 474

EP - 478

JO - SEG Technical Program Expanded Abstracts

JF - SEG Technical Program Expanded Abstracts

Y2 - 28 August 2023 through 1 September 2023

ER -

Revolutionizing seismic data compression: unlocking the power of stable diffusion neural networks

Abstract

Bibliographical note

ASJC Scopus subject areas

Access to Document

Fingerprint

Cite this