Michigan researchers have created a new 3D dataset that could transform how household robots understand their surroundings. Called 3D-GRAND, this AI-generated collection includes over 40,000 virtual rooms with millions of detailed descriptions, helping robots interpret 3D spaces using text. Early tests show it outperforms previous models by a wide margin. Find out more about this breakthrough and the team presenting their work at the upcoming CVPR Conference in the full article.
We use cookies to ensure you get the best experience on our website. By using this site, you accept our use of cookies.