🗺️What are the OVRMaps
OVRMaps are 3D Maps—digital twins—of the physical world, created by the OVER Community using smartphone cameras. Each map covers an average of 300 square meters and is created from 400 to 1000 multiview pictures. OVER processes these images with AI algorithms to generate a 3D model of the location. These maps are tied to their geographic coordinates, sharing the same H3 geographic reference system by Uber as OVRLands.
The OVER Mapping Process
OVER Mapping is a permissionless activity; anyone with the OVER App can map any location. Mappers can either own their map as an NFT or sell it to OVER through the Map2Earn program. This program incentivizes mappers with OVR Tokens for scanning high-traffic or high-value locations worldwide.
OVRMaps Applications
OVRMaps represent the 3D structure of physical locations and enable various applications in Physical AI:
Foundation Vision Models Training: The multiview pictures captured during mapping are used to train AI models that empower machines and robots understand and interact with the physical world.
XR Publishing and Browsing: Mapped locations can be explored remotely using real-time Gaussian Splat representations. This allows users to publish XR (Extended Reality) content remotely by using the 3D reconstruction as a reference. This also enables AR (Augmented Reality) content to be anchored to specific physical locations using the Visual Positioning System (VPS).
VPS Camera Pose Estimation Service: Once a location is mapped, the OVER VPS service can precisely estimate a camera’s position and orientation in 3D space with centimeter-level accuracy. This capability is crucial for machines and robots, allowing them to accurately determine their location and bearing within a 3D environment.
Last updated