Skip to content
Open access

Bidirectional Multi-Scale Cross-Modal Attention Network for SAR–Multispectral Semantic Segmentation

Oct 2026 · Remote Sensing · 0 citations · 38 references

TL;DR

Results confirm that BMCA effectively preserves fine-grained spatial details, promotes semantic consistency across modalities, and provides a robust and generalizable solution for feature-level SAR–MS fusion in remote-sensing semantic segmentation.

Abstract

Fusing synthetic aperture radar (SAR) and multispectral (MS) imagery for semantic segmentation remains challenging because of cross-modal semantic discrepancies, cloud-induced contamination, incomplete integration of complementary details, and the loss of spatial information during upsampling. To address these challenges, we propose BMCA, a feature-level SAR–MS fusion network for semantic segmentation and land-cover mapping. BMCA integrates bidirectional multiscale attention with cross-modal feature interaction. Specifically, the bidirectional multiscale attention module (DSM) captures deep interactions between the two modalities, refines hierarchical representations, and establishes cross-scale and cross-level dependencies, thereby enhancing both structural and semantic features. The cross-modal feature fusion module (CCF) adaptively integrates complementary information while suppressing modality-specific redundancy. Subsequently, the lightweight residual upsampling module (LRU) enhances spatial representations and recovers fine-grained details for pixel-level prediction. Extensive experiments were conducted on four remote-sensing datasets with diverse spatial resolutions and scene characteristics: SEN12MS, BigEarthNet, M3LEO, and MaRS-16M. On SEN12MS, BMCA achieves an F1-score of 86.76% and an IoU of 76.62%, surpassing the second-best method, CMX, by 1.13 and 1.75 percentage points, respectively. On the ultra-high-resolution MaRS-16M dataset, BMCA attains an F1-score of 97.48% and an IOU of 95.08%, exceeding CMX by 8.22 and 14.48 percentage points, respectively. Across the four datasets, BMCA obtains mean F1-score and IOU values of 92.58% and 86.42%, respectively, demonstrating consistent improvements over competing architectures. These results confirm that BMCA effectively preserves fine-grained spatial details, promotes semantic consistency across modalities, and provides a robust and generalizable solution for feature-level SAR–MS fusion in remote-sensing semantic segmentation.

Read PDF

Similar papers

Preprint Oct 2026

Cross-Scale Sparse Attention for Referring Segmentation in Remote Sensing

A Cross-Scale Sparse Attention Model (CSSAM), which aligns multiscale visual features with complex referring expressions through two key modules, and introduces M$^{3}RRSIS-D, a RRSIS dataset explicitly designed around three core reasoning dimensions: multi-target, reference-dependent multi-hop, and multi-granularity e...

Yu-Xi Li, Lu Si, Xia Liao et al. · 0 citations
Conference Sep 2026

Multi-source remote sensing data fusion based on attention-enhanced deep learning for intelligent land cover classification

The rapid growth of multi-source Earth observation data has introduced significant challenges for land cover classification. These challenges arise from two factors: cross-modal distribution gaps and spatial misalignment between optical imagery and synthetic aperture radar (SAR) data. This paper proposes AE-Net, an Att...

Yin-Hai Lu, Yue Sun · 0 citations
Open access 2026

LGCF-Net: Local–Global Cross-Modal Fusion With Residual-Gated Attention for Hyperspectral–LiDAR Land-Cover Classification

A local–global cross-modal fusion network (LGCF-Net), a hybrid fusion framework that explicitly models local–global feature interactions for HSI–LiDAR classification and achieves the best overall accuracy on MUUFL, Trento, and Augsburg, while attaining the highest average accuracy and Kappa on Houston.

M. Z. ur Rehman, S. Islam, Anwaar Ulhaq et al. · 0 citations
Open access Oct 2026

Dual-branch multimodal scale-interaction transformer for semi-supervised HSI-SAR classification

Deep learning has significantly advanced hyperspectral image (HSI) and synthetic aperture radar (SAR) joint classification, achieving promising land-cover mapping performance. However, most existing methods rely heavily on dense pixel-level annotations, which are costly and time-consuming to acquire in practical multim...

Ling-Ling Wu, Fei Gao, Fang Wang et al. · 0 citations
Open access 2026

MCESAN: A Modality Complementary Enhancement and Sparse Attention Network for Joint Hyperspectral and LiDAR Image Classification

Hyperspectral images (HSIs), with their hundreds of contiguous narrow bands, capture detailed physicochemical properties of land covers; LiDAR provides high-precision 3-D spatial and geometric structure, and the two modalities are highly complementary. However, differences in imaging mechanisms lead to insufficient cro...

Shu-Fang Xu, Wei-Wen Xu, Qi-Yuan Xue et al. · 0 citations
Preprint Nov 2026

Cross-Layer Refinement and Frequency-aware Fusion Network for Multi-Source Remote Sensing Image Classification

Multi-source remote sensing image classification benefits from complementary spectral and structural information. However, identifying discriminative cross-modal features remains challenging. This letter proposes a Cross-Layer Refinement and Frequency-aware Fusion Network (CRFNet) within a lightweight dual-branch archi...

Wen-Han Xiao, Chuan-Zheng Gong, Feng Gao et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.