NVIDIA, a graphics processor and mobile technologies company, is researching an AI technology that can convert 2D images into a 3D scene rapidly. The research involves a process called inverse rendering and applying it to a technology called neural radiance fields or NeRF. NVIDIA’s blog explains that inverse rendering is when an artificial intelligence is used to understand and mimic light behavior and uses several 2D images to create a 3D scene. NeRF technology works to fill in missing areas after estimating the light and color of a scene. In combining these into something labeled Instant NeRF, researchers have created the fastest process of its kind so far.
Using Instant NeRF requires many photos of the subject from different angles and information on the camera’s location. Using these images and data, the process will convert the resources into a 3D scene “within tens of milliseconds.”
David Luebke, the vice president for graphics research at NVIDIA predicts that “… Instant NeRF could be as important to 3D as digital cameras and JPEG compression have been to 2D photography — vastly increasing the speed, ease and reach of 3D capture and sharing”. His bold statement appears to have some weight as well, as in addition to the amazing speeds of Instant NeRF, it has a plethora of applications. The technology can be used to “create avatars or scenes for virtual worlds, to capture video conference participants and their environments in 3D, or to reconstruct scenes for 3D digital maps.” It seems that this technology might be useful both in the workforce and to hobbyists. Anyone from architects to animated filmmakers may benefit from this advancement.
It can be exciting to see how our world advances and changes. Sometimes existing technologies work great together and inspire new ones. In the future, what other technologies might be combined to improve our lives?