2020/04/15 by R. K. Gupta, Mubarak Shah, Gupta, Rohit +1 · 2 citations
Computer Science · Engineering · #Automated Road and Building Extraction #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #FOS: Electrical engineering #I.4.6 #Image and Video Processing (eess.IV) #Remote-Sensing Image Classification #Video Surveillance and Tracking Methods #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2004.07312
openalex publication_date 2020/04/15 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
Accurate and fine-grained information about the extent of damage to buildings\nis essential for directing Humanitarian Aid and Disaster Response (HADR)\noperations in the immediate aftermath of any natural calamity. In recent years,\nsatellite and UAV (drone) imagery has been used for this purpose, sometimes\naided by computer vision algorithms. Existing Computer Vision approaches for\nbuilding damage assessment typically rely on a two stage approach, consisting\nof building detection using an object detection model, followed by damage\nassessment through classification of the detected building tiles. These\nmulti-stage methods are not end-to-end trainable, and suffer from poor overall\nresults. We propose RescueNet, a unified model that can simultaneously segment\nbuildings and assess the damage levels to individual buildings and can be\ntrained end-toend. In order to to model the composite nature of this problem,\nwe propose a novel localization aware loss function, which consists of a Binary\nCross Entropy loss for building segmentation, and a foreground only selective\nCategorical Cross-Entropy loss for damage classification, and show significant\nimprovement over the widely used Cross-Entropy loss. RescueNet is tested on the\nlarge scale and diverse xBD dataset and achieves significantly better building\nsegmentation and damage classification performance than previous methods and\nachieves generalization across varied geographical regions and disaster types.\n