DRespNeT: A UAV Dataset and YOLOv8-DRN Model for Aerial Instance Segmentation of Building Access Points for Post-Earthquake Search-and-Rescue Missions

2025-08-21

AI Summary

Post-earthquake UAV aerial imagery presents significant challenges in accurately identifying and precisely annotating navigable entry points (e.g., doors, windows, gaps) and structural obstacles (e.g., collapsed buildings, multi-layer rubble), resulting in poor model generalization for disaster response. Method: We introduce DRN-DS—the first high-resolution, fine-grained instance segmentation dataset specifically designed for post-earthquake scenarios—containing 28 critical object classes with semantic distinctions for navigability. We further propose YOLOv8-DRN, an enhanced variant of YOLOv8-seg that integrates polygon-based annotation priors and a disaster-specific feature enhancement module. Contribution/Results: Evaluated on an RTX-4090 GPU, YOLOv8-DRN achieves 92.7% mAP₅₀ and 27 FPS, enabling real-time airborne instance segmentation. DRN-DS and YOLOv8-DRN collectively advance the accuracy of post-disaster path planning and enhance human–robot collaborative search-and-rescue efficiency.

Technology Category

• Computer Vision: Segmentation
• Intelligent Robots: Localization, Mapping, and Navigation
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Recent advancements in computer vision and deep learning have enhanced disaster-response capabilities, particularly in the rapid assessment of earthquake-affected urban environments. Timely identification of accessible entry points and structural obstacles is essential for effective search-and-rescue (SAR) operations. To address this need, we introduce DRespNeT, a high-resolution dataset specifically developed for aerial instance segmentation of post-earthquake structural environments. Unlike existing datasets, which rely heavily on satellite imagery or coarse semantic labeling, DRespNeT provides detailed polygon-level instance segmentation annotations derived from high-definition (1080p) aerial footage captured in disaster zones, including the 2023 Turkiye earthquake and other impacted regions. The dataset comprises 28 operationally critical classes, including structurally compromised buildings, access points such as doors, windows, and gaps, multiple debris levels, rescue personnel, vehicles, and civilian visibility. A distinctive feature of DRespNeT is its fine-grained annotation detail, enabling differentiation between accessible and obstructed areas, thereby improving operational planning and response efficiency. Performance evaluations using YOLO-based instance segmentation models, specifically YOLOv8-seg, demonstrate significant gains in real-time situational awareness and decision-making. Our optimized YOLOv8-DRN model achieves 92.7% mAP50 with an inference speed of 27 FPS on an RTX-4090 GPU for multi-target detection, meeting real-time operational requirements. The dataset and models support SAR teams and robotic systems, providing a foundation for enhancing human-robot collaboration, streamlining emergency response, and improving survivor outcomes.

Problem

• Aerial instance segmentation of building access points for post-earthquake SAR missions
• Identifying accessible entry points and structural obstacles in disaster zones
• Real-time detection of compromised structures and rescue targets using UAV imagery

Innovation

• High-resolution aerial dataset for detailed instance segmentation
• Optimized YOLOv8-DRN model achieves 92.7% mAP50
• Real-time inference at 27 FPS for operational requirements