Browsing by Author "Danoedoro, Projo"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Automatic landslide mapping with interpretable attention-based convolutional neural networks using remote sensing data
    (Association of Geoinformatics Technology, 2025-08-02) Mulabbi, Andrew; Danoedoro, Projo; Samodra, Guruh
    Landslide mapping plays a vital role in disaster management by providing essential information that can help decision making on mitigation and early warning strategies. However, existing automated methods often lack interpretability and miss crucial details, which limit their practical utility. This study addresses these limitations by introducing a novel Spatial Attention U-Net that leverages human visual attention to improve landslide detection and interpretability. Our proposed method integrates spatial attention modules throughout the U-Net's encoder and decoder paths, enabling the model to focus on critical image features for landslide identification. The model is trained and evaluated using a combination of high-resolution Pleiades RGB imagery, Brightness Index, and slope data. The model’s performance was evaluated using the F-1 score, precision, recall, and intersection over Union (IoU). The findings demonstrate that the Spatial Attention U-Net outperforms baseline models (U-Net, Squeeze-and-Excitation U-Net, and Channel-wise Attention), achieving F-1 scores of 73% and 79% on the testing and benchmark datasets, respectively. When applied to the inference/hold-out area, all the attention-based models outperformed the standard U-net, missing only three landslide events compared to five missed by the baseline model. Furthermore, the saliency maps reveal that the models focus on diverse regions of saliency, including edges, textures, tone, and brightness. The spatial attention U-net primarily highlights landslide edges (terrain discontinuities), while the baseline models use a mix of edges, texture, tone, and brightness. The results also indicate that dual-path attention does not lead to significant improvement in model accuracy. This approach offers a powerful tool for rapid and automated landslide mapping, indicating areas of saliency that can aid data annotation process by paying more attention to landslide object boundaries. The model interpretability further facilitates the creation of landslide inventories, especially in regions with limited ground truth data.
  • Thumbnail Image
    Item
    Integrating interpretable artificial neural networks, geomorphic plausibility and transfer learning for landslide susceptibility mapping: a case study of Pacitan, East Java
    (Springer Nature, 2025-10-27) Mulabbi, Andrew; Danoedoro, Projo; Samodra, Guruh
    Landslides are recurring natural hazards that continue to cause widespread damage to agricultural fields, vital infrastructures, and human lives globally. The impact is more severe in data-scarce regions where insufficient landslide inventory data limit accurate and timely landslide susceptibility assessment. This study integrates artificial neural networks, transfer learning, interpretable machine learning, and geomorphic plausibility to address the challenge of accurate landslide susceptibility mapping in data-scarce regions. Using the Pacitan regency as a case study, we assessed landslide susceptibility via artificial neural networks and transfer learning models. A neural network was first trained from a source area with a large dataset and then transferred through fine-tuning to another area with limited inventory data for mapping landslide susceptibility. A baseline model trained only on the limited target area data achieved an AUC test score of 0.83, the source-trained model had 0.88, while the transfer learning-based model achieved an AUC score of 0.97, confirming the improved performance and strong potential for transferability. Model interpretation using SHAP values, permutation importance, interaction strength, and partial dependence plots revealed slope, elevation, aspect, and distance to the stream as the most influential, while curvature and terrain indices were less impactful directly but contributed through interactions with other features. A qualitative assessment of geomorphic plausibility showed that the susceptibility patterns generally matched the behaviour of the terrain. Zones of high susceptibility were concentrated in the mid-slopes and at concave-convex transitions. Low susceptibility is observed in the steep rocky slope (> 40 °). These findings highlight the potential of interpretable machine learning, transfer learning, and geomorphic plausibility to produce physically meaningful susceptibility maps. This approach may be applied globally to produce rapid, interpretable, and generalisable landslide susceptibility maps in data-constrained regions, thereby contributing to more effective disaster mitigation and spatial planning.