학술논문
MATNet: Multilevel attention-based transformers for change detection in remote sensing images
Document Type
Article
Author
Source
In Image and Vision Computing November 2024 151
Subject
Language
ISSN
0262-8856
Abstract
Highlights •MATNet extracts multi-level features from bitemporal images.•The FEM can efficiently capture difference information.•The MAD with large-kernel attention fuse global and local information.•Tests on four datasets and ablation studies confirm MATNet's superior performance.