RMAU-NET: A Residual-Multihead-Attention U-Net Architecture for Landslide Segmentation and Detection from Remote Sensing Images

2025-07-15

AI Summary

Remote sensing-based landslide detection and pixel-wise segmentation remain challenging over large-scale, complex terrains (e.g., mountainous regions) due to small target sizes and intricate topographic textures. To address this, we propose Residual-Multihead-Attention U-Net (RMAU-Net), an end-to-end dual-task deep learning architecture that jointly optimizes landslide detection and segmentation by integrating residual connections with multi-head attention mechanisms. This design significantly enhances feature representation for small-scale landslides and heterogeneous terrain patterns. Evaluated on three public benchmarks—LandSlide4Sense, Bijie, and Nepal—RMAU-Net achieves F1 scores of 98.23/93.83 and mean Intersection-over-Union (mIoU) of 63.74/76.88, respectively, outperforming state-of-the-art methods. The results demonstrate RMAU-Net’s superior efficacy and robustness in operational remote sensing monitoring of landslide hazards.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Multi-instance/Multi-view Learning

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

In recent years, landslide disasters have reported frequently due to the extreme weather events of droughts, floods , storms, or the consequence of human activities such as deforestation, excessive exploitation of natural resources. However, automatically observing landslide is challenging due to the extremely large observing area and the rugged topography such as mountain or highland. This motivates us to propose an end-to-end deep-learning-based model which explores the remote sensing images for automatically observing landslide events. By considering remote sensing images as the input data, we can obtain free resource, observe large and rough terrains by time. To explore the remote sensing images, we proposed a novel neural network architecture which is for two tasks of landslide detection and landslide segmentation. We evaluated our proposed model on three different benchmark datasets of LandSlide4Sense, Bijie, and Nepal. By conducting extensive experiments, we achieve F1 scores of 98.23, 93.83 for the landslide detection task on LandSlide4Sense, Bijie datasets; mIoU scores of 63.74, 76.88 on the segmentation tasks regarding LandSlide4Sense, Nepal datasets. These experimental results prove potential to integrate our proposed model into real-life landslide observation systems.

Problem

• Automatically detect landslides from remote sensing images
• Segment landslide areas in large and rugged terrains
• Improve accuracy in landslide observation systems

Innovation

• Residual-Multihead-Attention U-Net architecture
• End-to-end deep learning for landslide detection
• Remote sensing image analysis for large terrains