DSBF-Net: A dual-branch split-attention network with spatial bilinear fusion for airborne remote sensing scene classification
Authors:
Open Access
- 1 January 2026
- journal article
- Published by American Institute of Mathematical Sciences (AIMS) in AIMS Geosciences
- Vol. 12 (3) , 847-889
Abstract
Airborne remote sensing scene classification is important for land-use surveys, urban planning, and environmental monitoring. However, large intra-class variation, strong inter-class similarity, and heterogeneous spatial structures make it difficult to coordinate representations extracted by different network pathways. To address this problem, this paper proposes a dual-branch split-attention network (DSBF-Net) with spatial bilinear fusion. EfficientNet and MobileNetV3 are employed as structurally heterogeneous backbones, Split-Attention is applied before and after fusion, and a spatial bilinear fusion module (SBFM) is introduced to align the two feature streams and model position-wise diagonal bilinear interactions. SBFM further incorporates signed square-root transformation, normalization, Coordinate Attention refinement, and residual aggregation. Experiments on the UC Merced Land-Use (UCM) dataset and Aerial Image Dataset (AID) show that DSBF-Net outperforms concatenation, addition, direct multiplication, and cross-attention fusion strategies, achieving classification accuracies of 99.07% and 97.58%, respectively. Component ablations further support the contribution of heterogeneous dual-branch learning, Split-Attention, and structured cross-branch fusion. The evaluation framework additionally examines backbone complementarity, hyperparameter sensitivity, corruption robustness, computational efficiency, interpretability, and limited-label behavior.Keywords
This publication has 36 references indexed in Scilit:
- FTransMamba: A multi-stage fusion transformer and mamba modeling for multimodal remote sensing scene understandingPattern Recognition, 2026
- HDC-Net:A multimodal remote sensing semantic segmentation network with hierarchical dual-stream fusion and cross-token interactionKnowledge-Based Systems, 2026
- ViCxLSTM: An extended Long Short-term Memory vision transformer for complex remote sensing scene classificationInternational Journal of Applied Earth Observation and Geoinformation, 2025
- Remote sensing scene classification with relation-aware dynamic graph neural networksEngineering Applications of Artificial Intelligence, 2025
- Improving semi-supervised remote sensing scene classification via Multilevel Feature Fusion and pseudo-labelingInternational Journal of Applied Earth Observation and Geoinformation, 2024
- Frequency and spatial based multi-layer context network (FSCNet) for remote sensing scene classificationInternational Journal of Applied Earth Observation and Geoinformation, 2024
- Few-shot remote sensing image scene classification: Recent advances, new baselines, and future trendsISPRS Journal of Photogrammetry and Remote Sensing, 2024
- Remote sensing scene classification using multi-domain sematic high-order networkImage and Vision Computing, 2024
- MSE-Net: A novel master–slave encoding network for remote sensing scene classificationEngineering Applications of Artificial Intelligence, 2024
- Remote sensing scene classification under scarcity of labelled samples—A survey of the state-of-the-artsComputers & Geosciences, 2023