Multi-Label Classification Framework for Hurricane Damage Assessment

2025-07-02

AI Summary

To address the limitation of single-label classification in post-hurricane damage assessment—namely, its inability to capture multiple damage types and severity levels simultaneously—this paper proposes a multi-label classification framework for aerial imagery. Methodologically, it employs ResNet as the backbone architecture, integrates a class-specific attention mechanism to enhance discriminative feature learning per damage category, and adopts a multi-label loss function to jointly predict damage types and their severity grades within a single image. Evaluated on the Hurricane Michael test set after training on the Rescuenet dataset, the framework achieves a mean Average Precision (mAP) of 90.23%, significantly outperforming existing baselines. This work is the first to introduce fine-grained, class-specific attention into multi-label disaster damage recognition, thereby improving both the accuracy and operational utility of post-disaster assessment. The paper has been accepted at the ASCE i3CE 2025 conference.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Hurricanes cause widespread destruction, resulting in diverse damage types and severities that require timely and accurate assessment for effective disaster response. While traditional single-label classification methods fall short of capturing the complexity of post-hurricane damage, this study introduces a novel multi-label classification framework for assessing damage using aerial imagery. The proposed approach integrates a feature extraction module based on ResNet and a class-specific attention mechanism to identify multiple damage types within a single image. Using the Rescuenet dataset from Hurricane Michael, the proposed method achieves a mean average precision of 90.23%, outperforming existing baseline methods. This framework enhances post-hurricane damage assessment, enabling more targeted and efficient disaster response and contributing to future strategies for disaster mitigation and resilience. This paper has been accepted at the ASCE International Conference on Computing in Civil Engineering (i3CE 2025), and the camera-ready version will appear in the official conference proceedings.

Problem

• Assessing diverse hurricane damage types and severities
• Overcoming limitations of single-label classification methods
• Improving accuracy in post-hurricane damage assessment

Innovation

• Multi-label classification for hurricane damage
• ResNet with attention mechanism
• Achieves 90.23% mean average precision