Machine-learning function for outputting camera viewpoints

A non-transitory computer readable storage medium having recorded thereon a computer program having instructions which, when executed by a computer, cause the computer to implement a method for determining a machine-learning function configured for taking an input 3D scene and for outputting one or more camera viewpoints each for generating a respective 2D rendering of the 3D scene, the method comprising:

• obtaining a library having 3D scenes and, for each 3D scene, one or more camera viewpoints each for generating a respective 2D rendering of the 3D scene;
• based on the library:
• forming a first dataset for training a first neural network configured for taking a first input representation of a 3D scene and for outputting information corresponding to a camera position in the 3D scene, and
• forming a second dataset for training a second neural network configured for taking a second input representation of a 3D scene and for outputting information corresponding to a camera orientation in the 3D scene;
• and
• training the first neural network based on the first dataset and training the second neural network based on the second dataset,
• each camera viewpoint outputted by the machine-learning function having a camera position corresponding to information outputted by the first neural network and a camera orientation corresponding to information outputted by the second neural network.

A system comprising:

• a processor coupled to a memory, the memory having recorded thereon a computer program for determining a machine-learning function configured for taking an input 3D scene and for outputting one or more camera viewpoints each for generating a respective 2D rendering of the 3D scene that when executed by the processor causes the processor to be configured to:
• obtain a library having 3D scenes and, for each 3D scene, one or more camera viewpoints each for generating a respective 2D rendering of the 3D scene;
• based on the library:
• form a first dataset for training a first neural network configured for taking a first input representation of a 3D scene and for outputting information corresponding to a camera position in the 3D scene, and
• form a second dataset for training a second neural network configured for taking a second input representation of a 3D scene and for outputting information corresponding to a camera orientation in the 3D scene;
• and
• train the first neural network based on the first dataset and training the second neural network based on the second dataset,
• each camera viewpoint outputted by the machine-learning function having a camera position corresponding to information outputted by the first neural network and a camera orientation corresponding to information outputted by the second neural network.