Explainable Quantum-Optimized Efficient Net Framework For Multi-Class Disaster Image Classification

Samiksha Bharti; Bappaditya Das

doi:10.69968/ijisem.2026v5i1198-208

Authors

Samiksha Bharti Department of Computer Science Engineering, Dr. B. C. Roy Engineering Collage, Durgapur, West Bangal, India
Bappaditya Das Department of Computer Science Engineering, Dr. B. C. Roy Engineering Collage, Durgapur, West Bangal, India

DOI:

https://doi.org/10.69968/ijisem.2026v5i1198-208

Keywords:

Disaster Detection, Explainable Artificial Intelligence, EfficientNetB0, Quantum Feature Selection, Deep Learning, Remote Sensing, Multi-Class Classification, Grad-CAM, Saliency Map, LIME

Abstract

Disaster detection needs to happen accurately and instantly because it helps to reduce human deaths and financial losses and environmental destruction. Satellite remote sensing and artificial intelligence advancements have created better systems which can automatically identify disasters through their analysis of extensive satellite imagery. Deep learning disaster detection systems function as black-box systems which restrict their ability to operate in emergency situations because decision-makers cannot understand their functioning. The research presents a deep learning system which combines explainability with quantum-enhanced efficiency to classify multiple disaster types through image analysis. The system uses EfficientNetB0 to extract features with high efficiency while using a quantum-inspired metaheuristic algorithm to select the best features and multiple Explainable Artificial Intelligence (XAI) methods which include Grad-CAM Saliency Maps and LIME. The proposed system is trained and evaluated on the Comprehensive Disaster Dataset (CDD) introduced by Niloy et al., comprising flood, wildfire, earthquake damage, landslide, urban fire, and non-disaster classes. The experimental assessment shows that the proposed system reaches 98.4% classification accuracy while maintaining high precision and recall and F1-score performance across all disaster categories. The model uses XAI techniques to provide visual and instance-level understanding of its predictions because the model needs to show how disaster features impact its outcomes. Comparative analysis indicates superior performance and improved explainability over recent deep learning–based disaster detection systems. The results confirm that combining quantum-based feature optimization with explainable deep learning enhances both predictive performance and operational trust, making the framework suitable for real-world disaster management and emergency response systems.

References

[1] Amit, Siti Nor Khuzaimah Binti, et al. "Analysis of satellite images for disaster detection." 2016 IEEE International geoscience and remote sensing symposium (IGARSS). IEEE, 2016.

[2] Janga, Bhargavi, et al. "A review of practical ai for remote sensing in earth sciences." Remote Sensing 15.16 (2023): 4112.

[3] Chuvieco, Emilio. Fundamentals of satellite remote sensing: An environmental approach. CRC press, 2020.

[4] Roy, P. S., M. D. Behera, and S. K. Srivastav. "Satellite remote sensing: sensors, applications and techniques." Proceedings of the National Academy of Sciences, India Section A: Physical Sciences 87.4 (2017): 465-472.

[5] Sahoh, Bukhoree, and Anant Choksuriwong. "The role of explainable Artificial Intelligence in high-stakes decision-making systems: a systematic review." Journal of Ambient Intelligence and Humanized Computing 14.6 (2023): 7827-7843.

[6] Mustafa, Ahmad M., et al. "Natural disasters detection using explainable deep learning." Intelligent Systems with Applications 23 (2024): 200430.

[7] Matin, Sahar S., and Biswajeet Pradhan. "Earthquake-induced building-damage mapping using Explainable AI (XAI)." Sensors 21.13 (2021): 4489.

[8] Liu, Shuxian, et al. "Evaluation of tropical cyclone disaster loss using machine learning algorithms with an explainable artificial intelligence approach." Sustainability 15.16 (2023): 12261.

[9] Raju, Akella S. Narasimha, et al. "GeoDisasterAINet: An Explainable Deep Ensemble Framework for Real-Time Urban and Rural Disaster Classification and Resilience." IEEE Access (2025).

[10] Collini, Enrico, et al. "Predicting and understanding landslide events with explainable AI." IEEE Access 10 (2022): 31175-31189.

[11] Hsiao, Po-Hsuan, et al. "Development of an explainable AI-based disaster casualty triage system." Computer Science and Information Systems 00 (2025): 35-35.

[12] Reddy, Chalamalla Nikhitha. "Explainable Artificial Intelligence (XAI) for Climate Hazard Assessment: Enhancing Predictive Accuracy and Transparency in Drought, Flood, and Landslide Modeling." IJSAT-International Journal on Science and Technology 16.1 (2025).

[13] Pradhan, Biswajeet, et al. "Spatial flood susceptibility mapping using an explainable artificial intelligence (XAI) model." Geoscience Frontiers 14.6 (2023): 101625.

[14] Cilli, Roberto, et al. "Explainable artificial intelligence (XAI) detects wildfire occurrence in the Mediterranean countries of Southern Europe." Scientific reports 12.1 (2022): 16349.

[15] Liu, Jia, et al. "Application of remote sensing and explainable artificial intelligence (XAI) for wildfire occurrence mapping in the mountainous region of southwest China." Remote Sensing 16.19 (2024): 3602.

[16] Choubin, Bahram, et al. "Explainable artificial intelligence (XAI) for interpreting predictive models and key variables in flood susceptibility." Results in Engineering (2025): 105976.

[17] Abdollahi, Arnick, and Biswajeet Pradhan. "Explainable artificial intelligence (XAI) for interpreting the contributing factors feed into the wildfire susceptibility prediction model." Science of the Total Environment 879 (2023): 163004.

[18] García-Tapia-Mateo, Paula, et al. "Explainable deep learning for early detection of natural disasters through social media text classification." Available at SSRN 5113748 (2025).

[19] Temenos, Anastasios, et al. "Novel insights in spatial epidemiology utilizing explainable AI (XAI) and remote sensing." Remote Sensing 14.13 (2022): 3074.

[20] Shin, Donghoon, et al. "Characterizing human explanation strategies to inform the design of explainable ai for building damage assessment." arXiv preprint arXiv:2111.02626 (2021).

[21] Alam, Gazi Mohammad Imdadul, et al. "Real-Time Detection of Forest Fires Using FireNet-CNN and Explainable AI Techniques." IEEE Access (2025)

[22] Mondal, Amit Kumar, et al. "A multi-model approach using XAI and anomaly detection to predict asteroid hazards." arXiv preprint arXiv:2503.15901 (2025).

[23] F. F. Niloy, A. B. S. Nayem, A. Sarker, O. Paul, M. A. Amin, A. A. Ali, M. I. Zaber, and A. K.M.M.Rahman,“A Novel Disaster Image Data-set and Characteristics Analysis using Attention Model,”Proc. 25th International Conference on Pattern Recognition (ICPR), pp. 6116–6122, 2021, IEEE.