Açıklama
Presenters: CVP Juan Lavista Ferres, Caleb Robinson, Cameron Birge, Kevin White
The Microsoft AI for Good Lab introduces HASTE, its open-source rapid building-damage assessment tool that turns post-disaster satellite, aerial, or drone imagery into building-by-building damage classifications in minutes. Through live demonstrations from real responses, the team shows both the interactive in-browser labeling workflow and the original semantic-segmentation approach, along with built-in validation tooling. The session also covers accuracy and validation, imagery sources, hard cases like flooding, and how results are activated and shared — with HASTE now fully open source for humanitarian and disaster-response organizations to run on their own infrastructure and build on via GitHub.
See more at https://www.microsoft.com/en-us/research/video/microsoft-ai-for-good-lab-introduction-to-haste/