Hurricanes cause significant damage to communities, necessitating rapid and accurate damage assessment to support timely disaster response. However, image-based deep learning models for hurricane-induced damage assessment face substantial challenges due to domain shifts across different hurricane events, and the restricted availability of labeled data for each disaster further complicates this task. In this study, we propose a novel domain-adaptive deep learning framework that mitigates the domain gap while requiring minimal labeled samples from the target domain. Our approach integrates a self-supervised learning (SSL) pretext task to enhance feature robustness and leverages a novel Bilateral Local Moran’s I module to improve spatial feature aggregation for damage localization. We evaluate our method using aerial datasets from Hurricanes Harvey, Matthew, and Michael. Experimental results demonstrate that our model achieves more than 5% improvement in damage classification accuracy over baseline methods. These findings highlight the potential of our approach for scalable and efficient hurricane damage assessment in real-world disaster scenarios.

Critical infrastructure (CI) plays a pivotal role in supporting daily life, encompassing vital sectors such as Healthcare, Transportation, and Power systems. Failure of CI can significantly impede everyday activities, making it essential to understand and address the interdependencies among these infrastructures. Failures can arise from both natural and manmade events, often leading to cascading effects across CI sectors. To mitigate capital loss and adversity during such failures, the efficient utilization of limited resources becomes crucial. This paper introduces a real-time decision-making system that incorporates local and global factors impacting CI and enables effective resource allocation. In this study, we focus on the failure of a specific CI sector, healthcare, caused by a flood event. To simulate this scenario, we create an environment that captures interdependencies between CI sectors, while generating flood events through randomly distributed water levels over time. To optimize decision-making, we employ a Reinforcement Learning (RL) based agent trained using Deep Q learning. The trained agent suggests critical decisions that enhance the utilization of limited resources, thereby extending the system's survivability. To provide a comprehensive view of the system's state and actions recommended by the learned agent at each time step (t), we develop a user interface. This interface displays the environment's states and facilitates the visualization of alternative CI protection strategies during catastrophic events like floods. Such simulation environments empower decision-makers with vital capabilities to make informed choices regarding resource allocation in critical scenarios.

Light Detection and Ranging (LiDAR) is an active remote sensing technique that uses pulsed lasers to sense the surrounding environment. It works on the principle of Time of Travel (ToT) for acquiring highly accurate and high spatial resolution 3D information about the surrounding environment. LiDAR data possesses immense potential for (near) real-time applications (e.g., forestry, disaster, border security, etc.) Real-time applications demand quick analysis of the data and cannot wait for the entire data pertaining to the area of interest to be captured before producing useful insights. Thus, 3D LiDAR points should be processed as and when they are captured in the form of a continuous stream. Due to the lack of prior knowledge about the (near) real-time data and the underlying distribution, in this work, unsupervised stream mining approaches have been adapted for the analysis of streaming geospatial 3D LiDAR data. By applying different unsupervised data mining (clustering) algorithms on a huge set of 3D LiDAR data points, we have evaluated the quality of clustering based on different evaluation metrics.

In a traditional Machine Learning setting, it is assumed that the training and test dataset belong to the same feature space and the same distribution. For practical real-world scenarios, this is not always true. The model which is trained on the source dataset might give poor results on the target dataset (during testing/inferencing). We address this challenge using Unsupervised Domain Adaptation which deals with the situation in which the network is trained on the labeled source domain data and unlabeled data belonging to a different but related target domain and it gives improved results when tested on the target domain data. Also, generating data considering two different but related domains into account is a challenge. In this work, we address this challenge by proposing a framework for developing a cross-domain Generative Adversarial Network (GAN) with the aim of achieving unsupervised domain adaptation which applies to LiDAR data appertaining to extreme events. The dataset used for the study is the pre- and post-earthquake LiDAR (Light Detection and Ranging) dataset belonging to the Kumamoto earthquake of 2016. A comprehensive comparative analysis is conducted in this study to examine the efficacy of urban damage classification, both with and without the integration of unsupervised domain adaptation. We envisage that our framework can be used for performing urban damage classification in an efficient manner with increased accuracy.

There has been an increase in the adoption of Linked Data and subsequently representing data in the form of knowledge graphs across a wide spectrum of domains. There has also been significant interest in the remote sensing community to publish Earth Observation data in the form of Linked Data. As the geospatial Linked Data cloud on the internet grows, there arises a need for efficient methods of exploratory analysis of such information-rich geospatial knowledge graphs. Knowledge graph representation of remote sensing scenes has proved to add significant value for effective mining of implicit information in addition to seamless integration with other data sources. This work is geared towards visual exploration of semantically enriched Remote Sensing Scene Knowledge Graphs (RSS-KGs). In this paper, we propose and implement an interactive web-based interface to visually explore and interact with RSS-KGs using Cesium. The proposed interface seeks to visualize the know ledge graph in the form of nodes and edges, mapped over the remote sensing scene consisting of different land use land cover regions and their inferred characteristics in addition to their spatial relationships with one another. It is envisaged that visualization in the form of nodes and edges would aid in visually validating the spatial relations in the knowledge graph, thus enhancing the understanding of the geospatial knowledge graph from the end user perspective. We demonstrate the efficacy of the interface through the visual exploration of an enriched geospatial knowledge graph of a remote sensing scene captured during an urban flood event.

Remote sensing lidar point cloud dataset embeds inherent 3D topological, topographical and complex geometrical information which possess immense potential in applications involving machine-understandable 3D perception. The lidar point clouds are unstructured, unlike images, and hence are challenging to process. In our work, we are exploring the possibility of deep learning-based question-answering on the lidar 3D point clouds. We are proposing a deep CNN-RNN parallel architecture to learn lidar point cloud features and word embedding from the questions and fuse them to form a feature mapping for generating answers. We have restricted our experiments for the urban domain and present preliminary results of binary question-answering (yes/no) using the urban lidar point clouds based on the perplexity, edit distance, evaluation loss, and sequence accuracy as the performance metrics. Our proposed hypothesis of lidar question-answering is the first attempt, to the best of our knowledge, and we envisage that our novel work could be a foundation in using lidar point clouds for enhanced 3D perception in an urban environment. We envisage that our proposed lidar question-answering could be extended for machine comprehension-based applications such as rendering lidar scene descriptions and content-based 3D scene retrieval.

In the early hours following the earthquake, supporting humanitarian actions like rescue operations and relief distribution is the primary objective of the rescue managers. The damage mapping can be performed using reliable data that can be obtained from high-resolution satellite imagery but obtaining satellite imagery can be challenging for some days post disaster due to revisit time. Considering the disaster response timing, Unmanned Aerial Vehicles (UAV) are used because ground transportation systems are ineffective due to road blockage. In this work, we make use of Light Detection and Ranging (LiDAR) 3D point cloud data obtained for Haiti Earthquake. The focus of our work is to develop and implement an approach for LiDAR data classification to enable Earthquake damage mapping and detection. This is obtained by running our deep learning network on NVIDIA Jetson Nano embedded supercomputing platform. This approach takes the advantage of embedded High-Performance computing and low power consumption capabilities of Jetson Nano which enhances the classification and promotes rapid response which is the key to manage post-disaster activities. Jetson Nano is a feasible option which provides a GPU architecture that is optimized for running energy-aware deep learning models and which generates the results in real or near-real time. We envisage that our work could be extended to perform near real-time classification of LiDAR point clouds in a post earthquake scenario.

Jetson TK1 is the first mobile processor from NVIDIA having similar features and architecture as that of a modern desktop GPU and still using low power from a mobile chip. Therefore, Jetson TK1 runs the same CUDA code (running on desktop GPU) with similar level of performance. Also, with the dawn of GPU technology, it has become possible to perform tasks (that are computationally intensive) in realtime or near-real time. In the agricultural domain, retrieving the biophysical parameters of the crop is important as it provides insights into the plant growth status. Inversion of the Radiative Transfer Model enables to obtain these parameters. However, such a process is highly computationally intensive. The focus of this work is to develop and implement an approach that takes the advantage of embedded High-Performance Computing (HPC) capability of Jetson TK1 to significantly improve the inversion process of a Radiative Transfer Model. The experimental results show that Jetson TK1 based biophysical parameters estimation gives significant speedup, which opens-up the possibility of having a Jetson based embedded platform for on-board biophysical parameters estimation in the future. In such a scenario, where there are constraints related to energy and power, Jetson TK1 can become a practicable option by providing a GPU based architecture for running energy-aware computationally intensive algorithms in parallel for processing the data, and generating the results in real-time or near-real time while taking care of the power usage.