A research team from Mohammed bin Zayed University for AI and other institutions, for the first time, was able to develop a multimedia model that supports the ability to process remote sensing and classify them semantically, in addition to linking each pixel in the image with categories or groups of specific things such as buildings, cars and semantic descriptions and this ability enables users to analyze images with high accuracy. The Geopixel model opens new horizons for understanding and analyzing the sensing images of remote accuracy, multiple and varying tasks in which users use multimedia models designed to understand text and analyze images, including classification of images, dividing them, and understanding the elements that make up them
source
