OUR NVIDIA in april technology Instant NeRF which quickly converts 2D images into a 3D scene, now the company has also introduced a new artificial intelligence 3D MoMa. This new technology from the company is a reverse rendering pipeline capable of reconstruct a 3D object using a series of still photographs.

According to Vice President of Graphic Research at NVIDIA, David Lübke, who does the work of converting still images into 3D models, “has long been the holy grail of combining computer vision and computer graphics.” To do this transformation job, the company’s innovation “uses state-of-the-art artificial intelligence and the raw processing power of NVIDIA GPUs to quickly create objects” that can be seamlessly imported, edited, and enhanced in existing editing tools.

Currently, the process of creating 3D objects is done manually or using photogrammetry methods, both of which take a significant amount of time. now with NVIDIA 3D MoMa time to create 3D models can be made fastercompatible with several existing tools and using only one NVIDIA GPU with Tensor Core.

O DPreview points out that new technology NVIDIA uses three main components to reconstruct an object, namely a 3D mesh model, materials, and lighting. The mesh works like the skeleton of the model, materials are 2D textures that are added to the skeleton, and finally an intended lighting model is created that users can modify later.

“In any traditional graphics engine, creators can easily change the material in the form created by NVIDIA 3D MoMa, as if they were wearing mesh with different clothes.”
– stated in the official note of NVIDIA.

NVIDIA has said it will submit a paper for NVIDIA 3D MoMa at the session of the conference on computer vision and pattern recognition, which will be held on June 22. In addition to this technology, the company said it was one of 38 papers with the company’s authors at the conference.

…..

Source: NVIDIA, DPreview