Method and apparatus for viewport shifting of non-real time 3d applications
AssigneeIntel Corporation
InventorsJoanna Douglas, Michal Taryma, Mario Garcia, Carlos Dominguez
Life of this patent
Estimating the term…
Abstract
Systems and methods for super sampling and viewport shifting of non-real time 3D applications are disclosed. In one embodiment, a graphics processing unit includes a processing resource to execute graphics commands to provide graphics for an application, a capture tool to capture the graphics commands, and a data generator to generate a dataset including at least one frame based on the captured graphics commands and to modify viewport settings for each frame of interest to generate a conditioned dataset.
Claims20
A graphics processing unit, comprising:
- memory to store graphics commands;
- and
- a streaming multiprocessor coupled to the memory, the streaming multiprocessor configured to:
- generate a first frame based on the graphics commands;
- generate a dataset based on the first frame and a plurality of replays of the graphics commands to generate a plurality of second frames having jittered viewpoints of the first frame, the dataset including a plurality of sample positions associated with modified viewpoints of the first frame;
- and
- train a machine learning model based on the dataset to enable generation of a third frame, the third frame a super sampled and anti-aliased frame having a higher resolution relative to the first frame.
A computer implemented method, comprising:
- storing graphics commands to memory of a graphics processor;
- generating a first frame based on the graphics commands;
- generating a dataset based on the first frame and a plurality of replays of the graphics commands to generate a plurality of second frames having jittered viewpoints of the first frame, the dataset including a plurality of sample positions associated with modified viewpoints of the first frame;
- and
- training a machine learning model based on the dataset to enable generation of a third frame, the third frame a super sampled and anti-aliased frame having a higher resolution relative to the first frame.
A data processing system comprising:
- a graphics processing unit comprising memory to store graphics commands and a streaming multiprocessor coupled to the memory, the graphics commands including commands associated with a graphics application programming interface (API), and the streaming multiprocessor configured to:
- generate a first frame based on the graphics commands;
- generate a dataset based on the first frame and a plurality of replays of the graphics commands to generate a plurality of second frames having jittered viewpoints of the first frame, the dataset including a plurality of sample positions associated with modified viewpoints of the first frame;
- and
- train a machine learning model based on the dataset to enable generation of a third frame, the third frame a super sampled and anti-aliased frame having a higher resolution relative to the first frame.
Dependencies are inferred from the claim text.