The SimCenter launched a project on Zooniverse, a crowd-sourcing website that leverages volunteer efforts of the general public to contribute to research projects. The purpose of SimCenter’s Zooniverse project, “Building Detective for Disaster Preparedness,” is to collect annotations of street-view images that can be utilized to train SimCenter’s research tool, Building Recognition Using Artificial Intelligence at Large Scale (BRAILS). BRAILS utilizes machine learning (ML) and deep learning (DL) to enhance building inventory databases, which are used in SimCenter’s regional analysis tools. Annotated images from the Zooniverse project will improve the accuracy of existing models and also train new models in the BRAILS tool.
The annotation process is simple and intuitive. A short tutorial can be viewed to better understand tasks, and a field guide provides annotation examples. A few moments of your time will have a high impact in improving computational simulation so that engineers can better understand the consequences of natural hazards on risk-prone buildings. Help spread the word!