The session recording will be archived on the site until June 25th, 2023

This session was streamed but not recorded

Curated Track: 

Description:

Scope:
Natural disasters are growing increasingly frequent and emerging in numerous forms, such as hurricanes, earthquakes, and wildfires. Some of them took places in densely populated regions, causing phenomenal devastations to human society. Additionally, other forms of disasters, such as health crisis, also bring detrimental impact to the society. More attentions have been paid by the government and researchers to address such issues and to alleviate the tension caused by disasters by enhancing the safety, resilience and sustainability of the communities.

Meanwhile, the technological leap in artificial intelligence has brought novel pathways for scholars to observe the formation and impact of disasters and unveil latent patterns through big geospatial data analysis. For example, supervised and semi-supervised methods have been widely applied in remote sensing imagery analysis to track and detect disasters; natural language processing applied in disaster related social media data analysis; neural network based regression models in disaster damage prediction. However, the application of advanced tools also raises challenges, namely the generalizability of topic specific deep learning models, the explainability of the black box models and the latent bias of the models that might be detrimental to social equity.

This session aims to accelerate the fusion of GeoAI and big data analytics for risk reduction. We welcome all latest works on the theory, methods, tools, as well as applications of GeoAI for disaster risk reductions with topics, including but not limited to:
• Geographic theory in deep learning approaches applied to disaster management and resilience assessment.
• AI aided (e.g., deep learning, computer vision, natural language processing) remote sensing and social media data mining in disaster management and resilience assessment.
• Time series modeling and prediction for disaster management and resilience assessment.
• Multi-modal data fusion (e.g., social media text, images, remote sensing images, street view images in disaster management and resilience assessment.
• Disaster vulnerability, resilience, and risk measurement and modeling.
• Large-scale spatiotemporal data processing, geo-computation and geo-visualization for disasters management.
• Novel geospatial datasets for geographic deep learning applied to disaster management and resilience assessment.
• Explainable AI and AI equity in disaster management.
• Disaster related graph representations.

Non-Presenting Participants