MSNet: A Multilevel Instance Segmentation Network for Natural Disaster Damage Assessment in Aerial Videos

  title={MSNet: A Multilevel Instance Segmentation Network for Natural Disaster Damage Assessment in Aerial Videos},
  author={Xiaoyu Zhu and Junwei Liang and Alexander Hauptmann},
  journal={2021 IEEE Winter Conference on Applications of Computer Vision (WACV)},
In this paper, we study the problem of efficiently assessing building damage after natural disasters like hurricanes, floods or fires, through aerial video analysis. We make two main contributions. The first contribution is a new dataset, consisting of user-generated aerial videos from social media with annotations of instance-level building damage masks. This provides the first benchmark for quantitative evaluation of models to assess building damage using aerial videos. The second… Expand

Figures and Tables from this paper

Comprehensive Semantic Segmentation on High Resolution UAV Imagery for Natural Disaster Damage Assessment
A large-scale hurricane Michael dataset is presented for visual perception in disaster scenarios, and state-of-the-art deep neural network models for semantic segmentation are analyzed. Expand
Comprehensive Semantic Segmentation on High Resolution Aerial Imagery for Natural Disaster Assessment
A large-scale hurricane Michael dataset is presented for visual perception in disaster scenarios, and state-of-the-art deep neural network models for semantic segmentation are analyzed. Expand
FloodNet: A High Resolution Aerial Imagery Dataset for Post Flood Scene Understanding
This paper presents a high resolution UAV imagery, FloodNet, captured after the hurricane Harvey and compares and contrast the performances of baseline methods for image classification, semantic segmentation, and visual question answering on this dataset. Expand
A Review on Automated Bone Age Measurement Based on Dental OPG Images
Bone Age Measurement (BAM) is a process of assessing skeletal maturity levels to measure one’s factual age. It has been used for clinical and medical investigation and forensic science. This paperExpand
Attention Based Semantic Segmentation on UAV Dataset for Natural Disaster Damage Assessment
A novel self-attention based semantic segmentation model is implemented on a high resolution UAV dataset and achieves Mean IoU score of around 88% on the test set, inspiring to use self-Attention schemes in natural disaster damage assessment which will save human lives and reduce economic losses. Expand
BDANet: Multiscale Convolutional Neural Network with Cross-directional Attention for Building Damage Assessment from Satellite Images
A novel two-stage convolutional neural network for Building Damage Assessment, called BDANet, which achieves state-of-the-art performance on a large-scale dataset – xBD. Expand
The application of UAV images in flood detection using image segmentation techniques
Received Sep 22, 2020 Revised Jun 26, 2021 Accepted Jul 3, 2021 The application of unmanned aerial vehicle (UAV) used to capture the images of the flood areas are becoming interest of mostExpand


Damage Assessment from Social Media Imagery Data During Disasters
This work analyzes images posted on social media platforms during natural disasters to determine the level of damage caused by the disasters and shows that the domain-specific fine-tuning of deep Convolutional Neural Networks (CNN) outperforms other state-of-the-art techniques such as Bag- of-Visual-Words (BoVW). Expand
Multi3Net: Segmenting Flooded Buildings via Fusion of Multiresolution, Multisensor, and Multitemporal Satellite Imagery
A novel approach for rapid segmentation of flooded buildings by fusing multiresolution, multisensor, and multitemporal satellite imagery in a convolutional neural network that allows for rapid and accurate post-disaster damage assessment and can be used by governments to better coordinate medium- and long-term financial assistance programs for affected areas. Expand
Benchmark Dataset for Automatic Damaged Building Detection from Post-Hurricane Remotely Sensed Imagery
A scalable framework for creating benchmark datasets of hurricane-damaged buildings and public sharing of the resulting benchmark datasets for Greater Houston area after Hurricane Harvey in 2017 are presented. Expand
Large-scale damage detection using satellite imagery
  • L. Gueguen, Raffay Hamid
  • Computer Science
  • 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2015
A semi-supervised learning framework for large-scale damage detection in satellite imagery that results in a ten-fold reduction in human annotation time at a minimal loss in detection accuracy compared to manual inspection is presented. Expand
Creating xBD: A Dataset for Assessing Building Damage from Satellite Imagery
xBD provides pre- and post-event multi-band satellite imagery from a variety of disaster events with building polygons, classification labels for damage types, ordinal labels of damage level, and corresponding satellite metadata, and will be the largest building damage assessment dataset to date. Expand
Convolutional Neural Networks for Disaster Images Retrieval
A late fusion technique is employed to jointly utilize visual and the additional information available in the form of meta-data for the retrieval of disaster image retrieval task in Mediaeval 2017 challenge on Multimedia and Satellite. Expand
Conditional Copulas for Change Detection in Heterogeneous Remote Sensing Images
A new preprocessing technique is presented in this paper to automatically highlight changes in multitemporal strongly heterogeneous remotely sensed images where the two acquisitions, before and after a given event, are significantly different, due, for instance, to different sensors, acquisition modalities, or climatic conditions. Expand
PolarMask: Single Shot Instance Segmentation With Polar Representation
  • Enze Xie, Pei Sun, +5 authors Ping Luo
  • Computer Science
  • 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2020
In this paper, we introduce an anchor-box free and single shot instance segmentation method, which is conceptually simple, fully convolutional and can be used by easily embedding it into mostExpand
Mask Scoring R-CNN
This paper proposes Mask Scoring R-CNN which contains a network block to learn the quality of the predicted instance masks and calibrates the misalignment between mask quality and mask score, and improves instance segmentation performance by prioritizing more accurate mask predictions during COCO AP evaluation. Expand
Earthquake damage mapping by using remotely sensed data: the Haiti case study
Classification results demonstrate that the simultaneous use of several change features from Earth observations can improve the damage estimation at object scale. Expand