A method for synthesizing novel 3D views of a scene by training a neural network to represent the scene as a volumetric density function.
Breakthrough in 3D reconstruction (2020).
Revolutionized how we capture reality.