Researchers at Allen Institute for Artificial Intelligence (AI2) have released an updated version of their OlmoEarth model, a family of Earth observation models. The new v1.1 release includes improvements to efficiency and accuracy. According to the developers, these changes will enable faster processing times without sacrificing performance. The updates were made using a combination of human expertise and automated testing. The OlmoEarth models are designed for tasks such as land cover classification and object detection in satellite imagery. --- Why it matters: This matters because more efficient Earth observation models can help researchers and organizations analyze large amounts of satellite data, which is crucial for applications like climate monitoring, natural disaster response, and urban planning. Source: https://huggingface.co/blog/allenai/olmoearth-v1-1