It’s tempting to think deep learning can just be thrown at a problem and will solve it. This is a concrete reminder that it's essential to understand your data distribution and its limitations, and to take steps to address those when you can.
CrowdAI examined the use of image segmentation models to improve the accuracy of temperature readings and other types of detection in thermal (IR) video.
CrowdAI announced today that it won an AFWERX Small Business Innovation and Research (SBIR) Phase II award from the US Air Force (AF) to modernize base building damage analysis.
CrowdAI used machine learning and human-in-the-loop review to produce a dataset of damaged buildings after the March 2020 tornadoes in Tennessee.
CrowdAI announced today that it won an AFWERX Small Business Innovation and Research (SBIR) Phase I grant from the US Air Force (AF) and the National Geospatial-Intelligence Agency (NGA).
CrowdAI deployed our deep learning models on NOAA imagery after Hurricane Dorian to understand the extent of the damage from flooding.