Back to results
US 12,737,843 B2Granted
Upsampling an image using one or more neural networks
AssigneeNVIDIA Corporation
InventorsShiqiu Liu, Robert Pottorff, Guilin Liu, Karan Sapra, Jon Barker, David Tarjan, Pekka Janis, Edvard Fagerholm, Lei Yang, Kevin Shih, Marco Salvi, Timo Roman, Andrew Tao, Bryan Catanzaro
Life of this patent
Estimating the term…
Abstract
Apparatuses, systems, and techniques are presented to generate images. In at least one embodiment, one or more neural networks are used to generate one or more images using one or more pixel weights determined based, at least in part, on one or more sub-pixel offset values.
Claims36
One or more processors, comprising:
- circuitry to use one or more neural networks to;
- determine a first blending weight and a second blending weight based, at least in part, on one or more upsampled images generated using one or more sub-pixel offsets associated with one or more low resolution images;
- perform, using the first blending weight, a first blending of pixel values of the one or more upsampled images with pixel values of one or more prior upsampled images to generate one or more intermediate image representations;
- perform, using the second blending weight, a second blending of the pixel values of the one or more upsampled images with pixel values of the one or more intermediate image representations;
- and
- generate one or more output images.
3 more elements
A method comprising:
- using one or more neural networks to;
- determine a first blending weight and a second blending weight based, at least in part, on one or more upsampled images generated using one or more sub-pixel offsets associated with one or more low resolution images;
- perform, using the first blending weight, a first blending of pixel values of the one or more upsampled images with pixel values of one or more prior upsampled images to generate one or more intermediate image representations;
- perform, using the second blending weight, a second blending of the pixel values of the one or more upsampled images with pixel values of the one or more intermediate image representations;
- and
- generate one or more output images.
3 more elements
A server, comprising:
- one or more processors to use one or more neural networks to;
- determine a first blending weight and a second blending weight based, at least in part, on one or more upsampled images generated using one or more sub-pixel offsets associated with one or more low resolution images;
- perform, using the first blending weight, a first blending of pixel values of the one or more upsampled images with pixel values of one or more prior upsampled images to generate one or more intermediate image representations;
- perform, using the second blending weight, a second blending of the pixel values of the one or more upsampled images with pixel values of the one or more intermediate image representations;
- and
- generate one or more output images.
3 more elements
A non-transitory machine-readable medium having stored thereon a set of instructions, which if performed by one or more processors, cause the one or more processors to at least use one or more neural networks to:
- determine a first blending weight and a second blending weight based, at least in part, on one or more upsampled images generated using one or more sub-pixel offsets associated with one or more low resolution images;
- perform, using the first blending weight, a first blending of pixel values of the one or more upsampled images with pixel values of one or more prior upsampled images to generate one or more intermediate image representations;
- perform, using the second blending weight, a second blending of the pixel values of the one or more upsampled images with pixel values of the one or more intermediate image representations;
- and
- generate one or more output images.
2 more elements
A video game system, comprising:
- one or more processors to use one or more neural networks to;
- determine a first blending weight and a second blending weight based, at least in part, on one or more upsampled images generated using one or more sub-pixel offsets associated with one or more low resolution images;
- perform, using the first blending weight, a first blending of pixel values of the one or more upsampled images with pixel values of one or more prior upsampled images to generate one or more intermediate image representations;
- perform, using the second blending weight, a second blending of the pixel values of the one or more upsampled images with pixel values of the one or more intermediate image representations;
- and
- generate one or more output images.
3 more elements
An autonomous vehicle, comprising:
- one or more processors to use one or more neural networks to;
- determine a first blending weight and a second blending weight based, at least in part, on one or more upsampled images generated using one or more sub-pixel offsets associated with one or more low resolution images;
- perform, using the first blending weight, a first blending of pixel values of the one or more upsampled images with pixel values of one or more prior upsampled images to generate one or more intermediate image representations;
- perform, using the second blending weight, a second blending of the pixel values of the one or more upsampled images with pixel values of the one or more intermediate image representations;
- and
- generate one or more output images.
3 more elements
Dependencies are inferred from the claim text.