September 10, 2025SensorsOpen Access

Deep Learning Image-Based Classification for Post-Earthquake Damage Level Prediction Using UAVs

View Full Paper

Ask AI

Bookmark

The MobileNetV3-Small model achieved a weighted average F-score of 0.93 in damage classification.
Experimental results showed a 58.8% reduction in FLOPs compared to ShuffleNetv2, enhancing efficiency.
The model was deployed on a Raspberry Pi 5, showcasing real-time performance in resource-constrained environments.
Fine-tuning improved accuracy by 4.5% over the baseline, emphasizing the model's robustness for practical use.

Discussion

No takes yet. Share an insight, caveat, or question.

This observational analysis predicts post-earthquake damage levels using UAV-captured images, indicating feasibility for SAR efforts.

Alsaaran et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark