programmers of Google developed RawNeRFan artificial intelligence program that is able to remove noise from images taken at nightwhat the experts called “see in the dark”.
According to a report on the website of computer today, this application is capable of combining photos taken from different angles and perspectives to combine them and reconstruct the scene.
RawNeRF works with Google’s Neural Radiance Fields, known simply as NeRF, which is a synthesis of views that create 3D renderings of a scene from multiple photos.
With the help of artificial intelligence, the program gives a second chance to RAW (raw) images, which are photographs taken in dark spaces and, when trying to illuminate them, what we do is make the mistake of filling them with noise (grains or bubbles). in the image), losing its quality.
The objective of RawNeRF
RawNeRF was created to remove that noise completely and give that dark image back a quality that it didn’t have before, allowing to see in detail all the elements of the photo.
Ben Mildenhall, one of the program developers, explained: “RawNeRF is not just a ‘denoiser’, which means that we can vary the position of the camera to see the scene from different angles”.
RawNeRF also reconstructs the scene and renders new areas in full color. The software is fed from a collection of up to 200 photos of the same scene, easily taken in seconds with today’s camera technology. You can manipulate focus, exposure, and tone mapping based on collected images.
For now, RawNeRF is in the testing process and could be available for both desktop and mobile devices, which is surely great news for photography lovers.