麻豆传媒

An artificial intelligence algorithm can transform still images into a high-resolution, explorable 3D world, with potential implications for film effects and virtual reality.

By feeding the neural network a selection of images of a scene and a rough 3D model of the scene created automatically using off-the-shelf software called , it is able to accurately visualise what the scene would look like from any viewpoint.

The neural network, developed by and colleagues at the University of Erlangen-Nuremberg in Germany, is different to previous systems because it is able to extract physical properties from still images.

鈥淲e can change the camera pose and therefore get a new view of the object,鈥� he says.

The system could technically create an explorable 3D world from just two images, but it wouldn鈥檛 be very accurate. 鈥淭he more images you have, the better the quality,鈥� says R眉ckert. 鈥淭he model cannot create stuff it hasn鈥檛 seen.鈥�

use between 300 and 350 images captured from different angles. R眉ckert hopes to improve the system by having it听simulate how light bounces off objects in the scene to reach the camera, which would mean fewer still images are needed for accurate 3D rendering.

鈥淯ntil now, creating photorealistic images from 3D reconstructions wasn鈥檛 fully automated and always had perceptible flaws,鈥� says , founder of New York-based company Abound Labs, who works on 3D capture software.

While Field points out the system still requires the input of accurate 3D data, and doesn鈥檛 yet work for moving objects, 鈥渢he rendering quality is unparalleled鈥�, he says. 鈥淚t鈥檚 proof that automated photorealism is possible.鈥�

Field believes the technology will be used for generating visual effects in films and virtual reality walkthroughs of locations. 鈥淚t鈥檚 going to accelerate the already-hot research field of machine learning-based rendering for computer generated imagery,鈥� he says.

Reference:

听

听