Abstract: The present disclosure relates to an apparatus and method for improving no-flash image quality by removing noise included in a no-flash image using a flash image. An apparatus for improving no-flash image quality using a flash image according to an example embodiment of the present disclosure includes an artificial neural network module which is trained by receiving an image pair of a flash image and a no-flash image and configured to output a convolutional kernel kc and a composite weight wci from a local flash image patch and a no-flash image patch, a convolution unit configured to convolve the flash image patch and the convolutional kernel to generate a consistent flash image patch, and a combining module configured to combine the no-flash image patch and the consistent flash image patch using the composite weight wci to produce a denoised image.

Abstract: The present disclosure relates to an apparatus and method for improving image quality by combining an unbiased image and a biased image using a James-Stein combiner. An apparatus for improving image quality using a James-Stein combiner according to the present disclosure includes an unbiased image buffer configured to output an unbiased image block, a biased image buffer configured to output a biased image block, a sample variance buffer configured to output a sample variance of the unbiased image, an artificial neural network model configured to derive a variance weight per pixel, a variance estimator configured to output a variance estimate of the unbiased image block, and a James-Stein combiner configured to locally combine the biased image block, the unbiased image block, and the variance estimate of the unbiased image block by applying a James-Stein combination equation.

Abstract: The present disclosure provides an apparatus for improving image quality including an image rendering module configured to generate a first data buffer and a second data buffer; an artificial neural network module configured to receive and learn the first data buffer and the second data buffer generated from the image rendering module; and a self-supervised loss calculation module that includes a first self-supervised loss function using pixel values of the first corrected image and the independent and correlated images of the second data buffer and a second self-supervised loss function using pixel values of the second corrected image and the independent and correlated images of the first data buffer, and a method for improving image quality.

Abstract: An apparatus for enhancing image quality includes a combining module configured to combine an independent image having pixels independent from each other and a correlated image including correlation information between pixels and an artificial intelligence (AI) apparatus configured to provide a weight used to combine the independent image and the correlated image.

Abstract: Techniques for selectively removing Monte Carlo (MC) noise from a geometric buffer (G-buffer). Embodiments identify the G-buffer for rendering an image of a three-dimensional scene from a viewpoint. Embodiments determine, for each of a plurality of pixels in the image being rendered, respective world position information based on the three-dimensional scene and a position and orientation of the viewpoint. A pre-filtering operation is then performed to selectively remove the MC noise from the G-buffer, based on the determined world position information for the plurality of pixels.

Abstract: Techniques for selectively removing Monte Carlo (MC) noise from a geometric buffer (G-buffer). Embodiments identify the G-buffer for rendering an image of a three-dimensional scene from a viewpoint. Embodiments determine, for each of a plurality of pixels in the image being rendered, respective world position information based on the three-dimensional scene and a position and orientation of the viewpoint. A pre-filtering operation is then performed to selectively remove the MC noise from the G-buffer, based on the determined world position information for the plurality of pixels.

Abstract: Techniques for selectively removing Monte Carlo (MC) noise from a geometric buffer (G-buffer). Embodiments identify the G-buffer for rendering an image of a three-dimensional scene from a viewpoint. Embodiments determine, for each of a plurality of pixels in the image being rendered, respective world position information based on the three-dimensional scene and a position and orientation of the viewpoint. A pre-filtering operation is then performed to selectively remove the MC noise from the G-buffer, based on the determined world position information for the plurality of pixels.

Abstract: Views of a virtual space may be presented based on predicted colors of individual pixels of individual frame images that depict the views of the virtual space. Predictive models may be assigned to individual pixels that predict individual pixel colors of individual pixels at individual time points. Individual models may be updated and/or reprojected to other pixels based on comparisons of the predicted pixel colors and colors specified by in a raster input signal.

Abstract: Techniques for selectively removing Monte Carlo (MC) noise from a geometric buffer (G-buffer). Embodiments identify the G-buffer for rendering an image of a three-dimensional scene from a viewpoint. Embodiments determine, for each of a plurality of pixels in the image being rendered, respective world position information based on the three-dimensional scene and a position and orientation of the viewpoint. A pre-filtering operation is then performed to selectively remove the MC noise from the G-buffer, based on the determined world position information for the plurality of pixels.

Abstract: Embodiments can provide adaptive image filtering. Under this approach, image quality can be enhanced by adjusting an approximation function to better adapt image signals in different parts of an image. Certain parts of the image may be enhanced using a certain approximation function while some other parts of the image may be enhanced using a different approximation function. In certain embodiments, the approximation function selected for a part of the image can be a polynomial function having a specific order. The specific polynomial order can be applied directly to obtain an estimated image value of the part of the image. In certain embodiments, the estimation of the reconstruction error can include iteratively estimating a bias term of the reconstruction error and a variance term of the reconstruction error.

Abstract: Views of a virtual space may be presented based on predicted colors of individual pixels of individual frame images that depict the views of the virtual space. Predictive models may be assigned to individual pixels that predict individual pixel colors of individual pixels at individual time points. Individual models may be updated and/or reprojected to other pixels based on comparisons of the predicted pixel colors and colors specified by in a raster input signal.

Abstract: Systems, methods and articles of manufacture for rendering an image. Embodiments include selecting a plurality of positions within the image and constructing a respective linear prediction model for each selected position. A respective prediction window is determined for each constructed linear prediction model. Additionally, embodiments render the image using the linear prediction models using the constructed linear prediction models, where at least one of the constructed linear prediction models is used to predict values for two or more of a plurality of pixels of the image, and where a value for at least one of the plurality of pixels is determined based on two or more of the constructed linear prediction models.

Abstract: Embodiments can provide adaptive image filtering. Under this approach, image quality can be enhanced by adjusting an approximation function to better adapt image signals in different parts of an image. Certain parts of the image may be enhanced using a certain approximation function while some other parts of the image may be enhanced using a different approximation function. In certain embodiments, the approximation function selected for a part of the image can be a polynomial function having a specific order. The specific polynomial order can be applied directly to obtain an estimated image value of the part of the image. In certain embodiments, the estimation of the reconstruction error can include iteratively estimating a bias term of the reconstruction error and a variance term of the reconstruction error.

Abstract: Systems, methods and articles of manufacture for rendering an image. Embodiments include selecting a plurality of positions within the image and constructing a respective linear prediction model for each selected position. A respective prediction window is determined for each constructed linear prediction model. Additionally, embodiments render the image using the linear prediction models using the constructed linear prediction models, where at least one of the constructed linear prediction models is used to predict values for two or more of a plurality of pixels of the image, and where a value for at least one of the plurality of pixels is determined based on two or more of the constructed linear prediction models.

Abstract: A method, system, and computer-readable storage medium are disclosed for adaptive sampling guided by multilateral filtering. A plurality of versions of a first image are generated. Each of the plurality of versions of the first image has a respective different resolution. A respective priority map is generated for each of the plurality of versions of the first image. Each respective priority map identifies a plurality of high-priority regions in a corresponding one of the plurality of versions of the first image. A second image is rendered based on the priority maps. The rendering comprises performing a ray-tracing process having a greater number of samples per pixel for the high-priority regions of the second image than for other regions of the second image.

Abstract: A method, system, and computer-readable storage medium are disclosed for adaptive sampling guided by multilateral filtering. A plurality of versions of a first image are generated. Each of the plurality of versions of the first image has a respective different resolution. A respective priority map is generated for each of the plurality of versions of the first image. Each respective priority map identifies a plurality of high-priority regions in a corresponding one of the plurality of versions of the first image. A second image is rendered based on the priority maps. The rendering comprises performing a ray-tracing process having a greater number of samples per pixel for the high-priority regions of the second image than for other regions of the second image.