Effective disaster management requires timely and accurate insights, yet traditional methods struggle to integrate multimodal data such as images, weather records, and textual reports. To address this, we propose DisasterNet-LLM, a specialized Large Language Model (LLM) designed for comprehensive disaster analysis. By leveraging advanced pretraining, cross-modal attention mechanisms, and adaptive transformers, DisasterNet-LLM excels in disaster classification. Experimental results demonstrate its superiority over state-of-the-art models, achieving higher accuracy of 89.5%, an F1 score of 88.0%, AUC of 0.92%, and BERTScore of 0.88% in multimodal disaster classification tasks.
@article{arxiv.2506.23462,
title = {Can We Predict the Unpredictable? Leveraging DisasterNet-LLM for Multimodal Disaster Classification},
author = {Manaswi Kulahara and Gautam Siddharth Kashyap and Nipun Joshi and Arpita Soni},
journal= {arXiv preprint arXiv:2506.23462},
year = {2025}
}
Comments
Accepted in the 2025 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2025), scheduled for 3 - 8 August 2025 in Brisbane, Australia