Masked Autoencoders for Sensor-Agnostic Remote Sensing Image Retrieval
Masked autoencoders can be effectively adapted to model inter-modal and intra-modal characteristics of multi-sensor remote sensing image archives for sensor-agnostic image retrieval.
Content-Adaptive Non-Local Convolution for Enhancing Remote Sensing Pansharpening
The proposed Content-Adaptive Non-Local Convolution (CANConv) module simultaneously incorporates spatial adaptability and non-local self-similarity to enhance the performance of remote sensing pansharpening.
Accurate Change Detection in High-Resolution Remote Sensing Imagery via Localization-Refinement Strategy
A novel change detection network, LRNet, is proposed based on a localization-then-refinement strategy to accurately discriminate change areas and their boundaries in high-resolution remote sensing imagery.
ChangeMamba: Efficient Spatio-Temporal Modeling for Remote Sensing Change Detection
The ChangeMamba architecture, based on the Mamba state space model, can efficiently model the global spatial context and spatio-temporal relationships to achieve accurate and efficient change detection in remote sensing images.
RS-Mamba: Efficient Global Modeling of Large Remote Sensing Images for Dense Prediction Tasks
Remote Sensing Mamba (RSM) is designed to efficiently model global features of very-high-resolution remote sensing images, enabling effective dense prediction tasks such as semantic segmentation and change detection.
Visual State Space Model for Efficient Semantic Segmentation of Remote Sensing Images
The proposed RS3Mamba model introduces a novel dual-branch architecture that incorporates a Visual State Space (VSS) auxiliary branch to provide additional global information, complementing the convolution-based main branch. A collaborative completion module is further introduced to effectively fuse the features from the two branches, enhancing the representation learning for remote sensing images.
Samba: A State Space Model-based Semantic Segmentation Framework for High-Resolution Remotely Sensed Images
Samba, a novel semantic segmentation framework built on the Mamba architecture, effectively captures global semantic information in high-resolution remotely sensed images with low computational complexity, outperforming state-of-the-art CNN and ViT-based methods.
Cross Modulation Transformer with Hybrid Loss for Enhancing Spatial and Spectral Fidelity in Remote Sensing Image Pansharpening
The proposed Cross Modulation Transformer (CMT) framework significantly advances pansharpening by introducing a novel modulation technique to effectively fuse high-resolution panchromatic and low-resolution multispectral images, while also employing a hybrid loss function that combines Fourier and wavelet transforms to capture both global and local image characteristics.
Efficient Global Feature Modeling for Remote Sensing Image Classification using State Space Model
RSMamba, an efficient global feature modeling methodology for remote sensing images based on the State Space Model (SSM), offers substantial advantages in representational capacity and efficiency, and is expected to serve as a feasible solution for handling large-scale remote sensing image interpretation.
© 2024 by Linnk AI