Abstract: Certain aspects of the present disclosure provide techniques for adaptive sampling for rendering using deep learning. This includes receiving, at a sampler in a rendering pipeline, a plurality of rendered pixel data, wherein the sampler includes a first machine learning (ML) model. It further includes generating a sampling map for the rendering pipeline using the first ML model and the plurality of rendered pixel data, including predicting a plurality of pixel values in the sampling map based on a generated distribution of pixel values. It further includes rendering an image using the sampler, the sampling map, and a denoiser in the rendering pipeline.

Abstract: A method and apparatus for rendering a plurality of objects. The method includes receiving ray differential information associated with a ray, determining a first resolution associated with a first surface patch based on at least information associated with the ray and the first surface patch, processing information associated with the first resolution and a database, and determining whether a first geometry representation associated with the first surface patch tessellated at the first resolution is present in the database. The database is associated with a plurality of predetermined resolutions, and the first resolution is equal to one of the plurality of predetermined resolutions. At least one of the plurality of predetermined resolutions includes a first tessellation rate associated with a first dimension and a second tessellation rate associated with a second dimension. The first tessellation rate is different from the second tessellation rate.

Abstract: A method and apparatus for rendering a plurality of objects. The method includes receiving ray differential information associated with a ray, determining a first ray differential based on at least information associated with the ray and a first surface patch, processing information associated with the first ray differential and the first surface patch, determining a first resolution based on at least information associated with the first ray differential and the first surface patch, and selecting a second resolution from the plurality of predetermined resolutions. Additionally, the method includes obtaining a first geometry representation associated with the second resolution.

Abstract: A method and apparatus for rendering a plurality of objects. The method includes receiving ray differential information associated with a ray, determining a first ray differential based on at least information associated with the ray and a first surface patch, and determining a first resolution based on at least information associated with the first ray differential and the first surface patch. Additionally, the method includes obtaining a first geometry representation based on at least information associated with the first surface patch and the first resolution, and obtaining a second geometry representation associated with a second surface patch tessellated at a second resolution. Moreover, the method includes determining whether there is at least one patch crack between the first geometry representation and the second geometry representation, and if there is the at least one patch crack, performing a stitching process to the first geometry representation and the second geometry representation.