NVIDIA 3D MoMa: New Methods to Improvise with 3D Objects

Dmitry Spodarets
Dmitry Spodarets

At the Computer Vision and Pattern Recognition Conference in New Orleans, the NVIDIA 3D MoMa inverse rendering pipeline was unveiled with a video about jazz and its homeland.

0:00
/

David Luebke, vice president of graphics research at NVIDIA, described the novelty as "a GPU-accelerated, differentiated component that, through its rendering pipeline, harnesses the mechanisms of modern AI and the raw computing power of the NVIDIA GPU to quickly create 3D objects."

Since game studios and other creators make such 3D objects using sophisticated photogrammetry techniques that require a lot of time and manual labor, NVIDIA 3D MoMa creates triangular mesh models within an hour on a single NVIDIA Tensor Core GPU.

Thus, graphics creators using NVIDIA 3D MoMa will one day gain the ability to improvise with 3D objects, importing, editing, and extending without the limitations of existing tools.

News