Convolutional sparse coding network for sparse seismic time-frequency representation

Seismic time-frequency (TF) transforms are essential tools in reservoir interpretation and signal processing, particularly for characterizing frequency variations in non-stationary seismic data. Recently, sparse TF transforms, which leverage sparse coding (SC), have gained significant attention in t...

Full description

Saved in:
Bibliographic Details
Main Authors: Qiansheng Wei, Zishuai Li, Haonan Feng, Yueying Jiang, Yang Yang, Zhiguo Wang
Format: Article
Language:English
Published: KeAi Communications Co. Ltd. 2025-06-01
Series:Artificial Intelligence in Geosciences
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2666544124000455
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849433426569986048
author Qiansheng Wei
Zishuai Li
Haonan Feng
Yueying Jiang
Yang Yang
Zhiguo Wang
author_facet Qiansheng Wei
Zishuai Li
Haonan Feng
Yueying Jiang
Yang Yang
Zhiguo Wang
author_sort Qiansheng Wei
collection DOAJ
description Seismic time-frequency (TF) transforms are essential tools in reservoir interpretation and signal processing, particularly for characterizing frequency variations in non-stationary seismic data. Recently, sparse TF transforms, which leverage sparse coding (SC), have gained significant attention in the geosciences due to their ability to achieve high TF resolution. However, the iterative approaches typically employed in sparse TF transforms are computationally intensive, making them impractical for real seismic data analysis. To address this issue, we propose an interpretable convolutional sparse coding (CSC) network to achieve high TF resolution. The proposed model is generated based on the traditional short-time Fourier transform (STFT) transform and a modified UNet, named ULISTANet. In this design, we replace the conventional convolutional layers of the UNet with learnable iterative shrinkage thresholding algorithm (LISTA) blocks, a specialized form of CSC. The LISTA block, which evolves from the traditional iterative shrinkage thresholding algorithm (ISTA), is optimized for extracting sparse features more effectively. Furthermore, we create a synthetic dataset featuring complex frequency-modulated signals to train ULISTANet. Finally, the proposed method's performance is subsequently validated using both synthetic and field data, demonstrating its potential for enhanced seismic data analysis.
format Article
id doaj-art-558d982f2afe4a14ad8a7445ea78bd9c
institution Kabale University
issn 2666-5441
language English
publishDate 2025-06-01
publisher KeAi Communications Co. Ltd.
record_format Article
series Artificial Intelligence in Geosciences
spelling doaj-art-558d982f2afe4a14ad8a7445ea78bd9c2025-08-20T03:27:02ZengKeAi Communications Co. Ltd.Artificial Intelligence in Geosciences2666-54412025-06-016110010410.1016/j.aiig.2024.100104Convolutional sparse coding network for sparse seismic time-frequency representationQiansheng Wei0Zishuai Li1Haonan Feng2Yueying Jiang3Yang Yang4Zhiguo Wang5The Third Gas Production Plant of PetroChina Changqing Oilfield Branch, Wushenqi, Inner Mongolia, 017000, PR ChinaSchool of Mathematics and Statistics, Xi'an Jiaotong University, Xi'an, Shaanxi, 710049, PR ChinaThe Third Gas Production Plant of PetroChina Changqing Oilfield Branch, Wushenqi, Inner Mongolia, 017000, PR ChinaSchool of Mathematics and Statistics, Xi'an Jiaotong University, Xi'an, Shaanxi, 710049, PR ChinaSchool of Information and Communications Engineering, Xi'an Jiaotong University, Xi'an, Shaanxi, 710049, PR ChinaSchool of Mathematics and Statistics, Xi'an Jiaotong University, Xi'an, Shaanxi, 710049, PR China; Corresponding author.Seismic time-frequency (TF) transforms are essential tools in reservoir interpretation and signal processing, particularly for characterizing frequency variations in non-stationary seismic data. Recently, sparse TF transforms, which leverage sparse coding (SC), have gained significant attention in the geosciences due to their ability to achieve high TF resolution. However, the iterative approaches typically employed in sparse TF transforms are computationally intensive, making them impractical for real seismic data analysis. To address this issue, we propose an interpretable convolutional sparse coding (CSC) network to achieve high TF resolution. The proposed model is generated based on the traditional short-time Fourier transform (STFT) transform and a modified UNet, named ULISTANet. In this design, we replace the conventional convolutional layers of the UNet with learnable iterative shrinkage thresholding algorithm (LISTA) blocks, a specialized form of CSC. The LISTA block, which evolves from the traditional iterative shrinkage thresholding algorithm (ISTA), is optimized for extracting sparse features more effectively. Furthermore, we create a synthetic dataset featuring complex frequency-modulated signals to train ULISTANet. Finally, the proposed method's performance is subsequently validated using both synthetic and field data, demonstrating its potential for enhanced seismic data analysis.http://www.sciencedirect.com/science/article/pii/S2666544124000455Time-frequency transformIteration shrinkage threshold algorithmDeep learningUNet
spellingShingle Qiansheng Wei
Zishuai Li
Haonan Feng
Yueying Jiang
Yang Yang
Zhiguo Wang
Convolutional sparse coding network for sparse seismic time-frequency representation
Artificial Intelligence in Geosciences
Time-frequency transform
Iteration shrinkage threshold algorithm
Deep learning
UNet
title Convolutional sparse coding network for sparse seismic time-frequency representation
title_full Convolutional sparse coding network for sparse seismic time-frequency representation
title_fullStr Convolutional sparse coding network for sparse seismic time-frequency representation
title_full_unstemmed Convolutional sparse coding network for sparse seismic time-frequency representation
title_short Convolutional sparse coding network for sparse seismic time-frequency representation
title_sort convolutional sparse coding network for sparse seismic time frequency representation
topic Time-frequency transform
Iteration shrinkage threshold algorithm
Deep learning
UNet
url http://www.sciencedirect.com/science/article/pii/S2666544124000455
work_keys_str_mv AT qianshengwei convolutionalsparsecodingnetworkforsparseseismictimefrequencyrepresentation
AT zishuaili convolutionalsparsecodingnetworkforsparseseismictimefrequencyrepresentation
AT haonanfeng convolutionalsparsecodingnetworkforsparseseismictimefrequencyrepresentation
AT yueyingjiang convolutionalsparsecodingnetworkforsparseseismictimefrequencyrepresentation
AT yangyang convolutionalsparsecodingnetworkforsparseseismictimefrequencyrepresentation
AT zhiguowang convolutionalsparsecodingnetworkforsparseseismictimefrequencyrepresentation