New technology from NVIDIA turns photos into three-dimensional objects

Is called Neural Radiance FieldO NeRFand provides for the training of some specific algorithms of artificial intelligence to transform the classics two-dimensional photos in three-dimensional objects. NeRF adds all those details that the photos have not captured to fill in the empty spaces and give three-dimensional thickness to the photographed subjects.

NVIDIA: a new way to use artificial intelligence

It only takes a little forethought to complete the process. To train the AI ​​model, defined Instant NeRF, you need to provide photos from various angles, along with where the camera was at the time the shot was taken. In addition, for photos with multiple subjects, movement should be kept to a minimum, otherwise the result may be blurry.

The technique is based on a hardware encoding that NVIDIA defines “multi-resolution hash grid”, designed to run efficiently by its GPUs. It also works with traditional GPUs, but those equipped with Tensor Core, that is specific hardware for the calculations related to artificial intelligence, will produce the results in less time. According to NVIDIA, if it normally takes hours to perform this type of rendering, the processing work is completed in seconds with your hardware.

The training part takes longer, while during the rendering phase the NVIDIA GPUs are able to infer missing data and reconstruct realistic three-dimensional scenes in seconds.

NVIDIA’s new technology could lead to significant advancements in various fields, from autonomous driving up to video games. In the case of autonomous driving, it could be used to help on-board computers in vehicles better understand the dimensions and shapes of real-world objects. Me too’architecture it could be an area of ​​use, generating 3D models of real environments that can be modified.

Gift ideas, why waste time and risk making mistakes?

GIVE A GOOD AMAZON!