Matterport and Facebook AI Research Collaborate to Unveil the World's Largest Dataset of 3D Spaces for Academic Research

HM3D is now available for free academic and non-commercial research. Researchers can utilize it in conjunction with FAIR's Habitat simulator to train embodied agents at scales, such as home robots and AI assistants.

FREMONT, CA: Matterport, the spatial data company leading the digital transformation of the built world, which has entered into a definitive agreement to enter into a business combination with Gores Holdings VI, has declared a collaboration with Facebook AI Research (FAIR) through which it will make the largest-ever dataset of 3D indoor spaces available exclusively for academic, non-commercial uses. The Habitat-Matterport 3D Research Dataset (HM3D) is a first-of-its-kind collection of 1,000 high-resolution Matterport digital twins of residential, commercial, as well as civic areas produced precisely from real-world surroundings. HM3D will play an important part in the advancement of embodied AI research, which aims to teach robots and virtual AI helpers how to understand and interact with the intricacies of the physical environment.

"Until now, this rich spatial data has been glaringly absent in the field, so HM3D has the potential to change the landscape of embodied AI and 3D computer vision," stated Dhruv Batra, Research Scientist at Facebook AI Research. "Our hope is that the 3D dataset brings researchers closer to building intelligent machines, to do for embodied AI what pioneers before us did for 2D computer vision and other areas of AI."

HM3D is now available for free academic and non-commercial research. Researchers can utilize it in conjunction with FAIR's Habitat simulator to train embodied agents at scales, such as home robots and AI assistants. HM3D is a critical step toward assisting these agents in navigating real-world environments and better understanding the variations of spaces such as bedrooms, bathrooms, kitchens, corridors, and the many arrangements of those rooms within each structure. It can also help robots recognize how objects in rooms are usually arranged so that instructions are accurately comprehended. One day, this discovery could be employed in manufacturing applications such as robots retrieving medicine from a bedside nightstand or augmented reality glasses that assist individuals in remembering where they left their keys.