Swin-ReshoUnet: A Seismic Profile Signal Reconstruction Method Integrating Hierarchical Convolution, ORCA Attention, and Residual Channel Attention Mechanism

Swin-ReshoUnet: A Seismic Profile Signal Reconstruction Method Integrating Hierarchical Convolution, ORCA Attention, and Residual Channel Attention Mechanism

This study proposes a Swin-ReshoUnet architecture with a three-level enhancement mechanism to address inefficiencies in multi-scale feature extraction and gradient degradation in deep networks for high-precision seismic exploration. The encoder uses a hierarchical convolution module to build a multi...

Full description

Saved in:
Bibliographic Details
Main Authors: Jie Rao, Mingju Chen, Xiaofei Song, Chen Xie, Xueyang Duan, Xiao Hu, Senyuan Li, Xingyue Zhang
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Applied Sciences
Subjects:
Swin-ReshoUnet
three-level enhancement
hierarchical convolution
ORCA attention
residual channel attention
seismic profile signal reconstruction
Online Access:https://www.mdpi.com/2076-3417/15/15/8332
Tags: Add Tag
No Tags, Be the first to tag this record!

Internet

https://www.mdpi.com/2076-3417/15/15/8332

Similar Items