Deep Learning for Earthquake Disaster Assessment: Objects, Data, Models, Stages, Challenges, and Opportunities

Remote Sensing, Год журнала: 2023, Номер 15(16), С. 4098 - 4098

Опубликована: Авг. 21, 2023

Earthquake Disaster Assessment (EDA) plays a critical role in earthquake disaster prevention, evacuation, and rescue efforts. Deep learning (DL), which boasts advantages image processing, signal recognition, object detection, has facilitated scientific research EDA. This paper analyses 204 articles through systematic literature review to investigate the status quo, development, challenges of DL for The first examines distribution characteristics trends two categories EDA assessment objects, including earthquakes secondary disasters as buildings, infrastructure, areas physical objects. Next, this study application distribution, advantages, disadvantages three types data (remote sensing data, seismic social media data) mainly involved these studies. Furthermore, identifies six commonly used models EDA, convolutional neural network (CNN), multi-layer perceptron (MLP), recurrent (RNN), generative adversarial (GAN), transfer (TL), hybrid models. also systematically details at different times (i.e., pre-earthquake stage, during-earthquake post-earthquake multi-stage). We find that most extensive field involves using CNNs classification detect assess building damage resulting from earthquakes. Finally, discusses related training models, opportunities new sources, multimodal DL, concepts. provides valuable references scholars practitioners fields.

Язык: Английский

---

Artificial intelligence-Enabled deep learning model for multimodal biometric fusion

Multimedia Tools and Applications, Год журнала: 2024, Номер 83(33), С. 80105 - 80128

Опубликована: Фев. 8, 2024

Язык: Английский

---

Improved deep learning method and high-resolution reanalysis model-based intelligent marine navigation

Frontiers in Marine Science, Год журнала: 2025, Номер 12

Опубликована: Апрель 14, 2025

Large-scale weather forecasting is critical for ensuring maritime safety and optimizing transoceanic voyages. However, sparse meteorological data, incomplete forecasts, unreliable communication hinder accurate, high-resolution wind system predictions. This study addresses these challenges to enhance dynamic voyage planning intelligent ship navigation. We propose IPCA-MHA-DSRU-Net, a novel deep learning model integrating incremental principal component analysis (IPCA) with spatial-temporal depthwise separable U-Net. Key components include: (1) IPCA preprocessing reduce dimensionality noise in 2D field data; (2) depthwise-separable convolution (DSC) blocks minimize parameters computational costs; (3) multi-head attention (MHA) residual mechanisms improve feature extraction prediction accuracy. The framework optimized real-time onboard deployment under constraints. achieves high accuracy predictions, validated through reanalysis datasets. Experiments demonstrated enhanced path efficiency robustness oceanic conditions. IPCA-MHA-DSRU-Net balances accuracy, making it viable resource-limited ships. application provides promising alternative large-scale data.

Язык: Английский

---

Applications of artificial intelligence technologies in water environments: From basic techniques to novel tiny machine learning systems

Process Safety and Environmental Protection, Год журнала: 2023, Номер 180, С. 10 - 22

Опубликована: Сен. 30, 2023

Язык: Английский

---

Cascading hazards from two recent glacial lake outburst floods in the Nyainqêntanglha range, Tibetan Plateau

Journal of Hydrology, Год журнала: 2023, Номер 626, С. 130155 - 130155

Опубликована: Сен. 14, 2023

Язык: Английский

---