Abstract: In some aspects, the present disclosure provides a method for high dynamic range (HDR) video rotation. The method includes receiving, by a display processor, an indication that a frame rotation animation process for video playback has been initiated, the display processor comprising a display processor pipeline. The method also includes determining whether the video playback is an HDR format or another format. In response to the determination and receiving the indication: bypassing a loading of the frame rotation animation into a first portion of the display processor pipeline, and loading the frame rotation animation into a second portion of the display processor pipeline if the video playback is in an HDR format.

Abstract: Systems, apparatuses, and methods for loading multiple primitives per thread in a graphics pipeline are disclosed. A system includes a graphics pipeline frontend with a geometry engine, shader processor input (SPI), and a plurality of compute units. The geometry engine generates primitives which are accumulated by the SPI into primitive groups. While accumulating primitives, the SPI tracks the number of vertices and primitives per group. The SPI determines wavefront boundaries based on mapping a single vertex to each thread of the wavefront while allowing more than one primitive per thread. The SPI launches wavefronts with one vertex per thread and potentially multiple primitives per thread. The compute units execute a vertex phase and a multi-cycle primitive phase for wavefronts with multiple primitives per thread.

Abstract: Systems and techniques are provided for enhancing operations of a ray tracing processor. For instance, a process can include obtaining one or more nodes of an acceleration data structure. Each node of the one or more nodes includes the same number of bytes. The node(s) can be stored in a cache associated with a ray tracing processor. Each of the stored node(s) are cache line-aligned with the cache associated with the ray tracing processor. A first stored node of the stored node(s) can be provided to the ray tracing processor and processed by the ray tracing processor during a first clock cycle of the ray tracing processor. A second stored node of the stored node(s) can be provided to the ray tracing processor and processed by the ray tracing processor during a second clock cycle of the ray tracing processor.

Abstract: The present disclosure describes techniques of rendering images using explicit object representation via rays tracing volume density aggregation. The techniques comprise reconstructing an object into a plurality of Gaussian ellipsoids; determining a volume density of each of the plurality of Gaussian ellipsoids along each of a plurality of viewing rays; determining a weight of each of the plurality of Gaussian ellipsoids based on the volume density; and synthesizing an image of the object using the determined weight on each pixel of the image to interpolate attributes of each of the plurality of Gaussian ellipsoids.

Abstract: A system and method enabling per-user-optimized computing, rendering, and provisioning within virtual worlds. The system comprises a server including memory and at least one processor, the memory storing a persistent virtual world system comprising a data structure in which at least one virtual replica of at least one corresponding real object is represented, and a computing optimization platform configured to store and provide rules for optimizing the computing, rendering and data provisioning to users via user devices. A plurality of connected devices connected to the server via a network provide multi-source data, user input, or combinations thereof, to the persistent virtual world system, updating the virtual replicas. The server retrieves user location, viewing position and orientation from the one or more user devices to determine a user interaction radius, thereby optimizing via the computing optimization platform the relevant computing, rendering and provisioning for the one or more user devices.

Abstract: A method and apparatus for simulating clothes receive a user setting for an area in which shirring is to be expressed in a two-dimensional (2D) pattern of clothes modeled with a mesh including a plurality of polygons, obtain a first line and a second line parallel to the first line from the area based on the user setting, generate inner lines that are perpendicular to at least one of the first line and the second line of the area between the first line and the second line and that have regular intervals, align polygons included in the area of the 2D pattern based on the inner lines, and perform a draping simulation of three-dimensional (3D) clothes corresponding to the 2D pattern in which the shirring is expressed by the aligned polygons.

Abstract: The present disclosure relates to methods and apparatus for graphics processing. The apparatus may identify at least one mesh associated with at least one frame. The apparatus may also divide the at least one mesh into a plurality of groups of primitives, each of the plurality of groups of primitives including at least one primitive and a plurality of vertices. The apparatus may also compress the plurality of groups of primitives into a plurality of groups of compressed primitives, the plurality of groups of compressed primitives being associated with random access. Additionally, the apparatus may decompress the plurality of groups of compressed primitives, at least one first group of the plurality of groups of compressed primitives being decompressed in parallel with at least one second group of the plurality of groups of compressed primitives.

Abstract: In one embodiment, a method includes identifying a portion of an object representation in a virtual scene to be displayed by a set of pixels of a display, accessing a rendering pattern associated with the set of pixels, wherein the rendering pattern is associated with the set of pixels based on a position of the set of pixels relative to a foveal region, accessing a texture array associated with the portion of the object representation, determining, according to the rendering pattern, first pixel values for a first subset of the set of pixels by sampling the texture array, determining, according to the rendering pattern, second pixel values for a second subset of the set of pixels using the first pixel values, and sending the first pixel values and the second pixel values to the display.

Abstract: Systems and methods are disclosed for the generation and interactive display of three-dimensional (3D) models that accurately represent buildings and other large, complex objects possessing both internal and external features. More specifically, an exterior mesh can be generated that captures all the external features of the building. An interior mesh can also be generated that captures all the internal features of the building. The exterior mesh and the interior mesh may be layered and combined using geographical coordinates (e.g., global positioning system (GPS) data) in order to create a combined mesh that can be used to render a 3D model of the object. Any images taken of external or internal features may also be appended to the 3D model or even overlaid on the 3D model (like textures), based on the location and directional view associated with those images.

Abstract: A positioning method includes clustering points in a first point cloud through multi-clustering to obtain a target point cloud, where the target point cloud represents a feature of a target object, and the first point cloud includes the target point cloud and a point cloud that represents a feature of an interfering object; and determining a position of the target object based on the target point cloud.

Abstract: Systems, methods, and/or computer-readable media described herein provide technical solutions to aid in the color matching of dental restorations. These systems, methods and/or computer readable media may receive a three-dimensional (3D) model of a tooth, where the 3D model includes a plurality of voxels, set each voxel to a starting optical property, and form a 3D visible light model from the 3D model by iteratively adjusting the optical property of each voxel based on optical properties measured from a scan of the tooth, where the 3D visible light model provides information to generate a color matched dental restoration.

Abstract: The present disclosure provides systems and methods for generating photorealistic image simulation data with geometry-aware composition for testing autonomous vehicles. In particular, aspects of the present disclosure can involve the intake of data on an environment and output of augmented data on the environment with the photorealistic addition of an object. As one example, data on the driving experiences of a self-driving vehicle can be augmented to add another vehicle into the collected environment data. The augmented data may then be used to test safety features of software for a self-driving vehicle.

Abstract: A method of operation of a texturing/shading unit in a GPU pipeline is used for efficient convolution operations. The method uses texture hardware to collectively fetch all the texels required to calculate properties for a group of output pixels without any duplication. The method then bypasses bilinear filter hardware in the texture hardware and passes the fetched and unfiltered texel data from the texture hardware unit to shader hardware in the texturing/shading unit. The shader hardware uses the fetched texel data to perform a plurality of convolution operations to calculate the properties of each of the output pixel.

Abstract: One embodiment of a virtual reality apparatus comprises: a graphics processing engine comprising a plurality of graphics processing stages, the graphics processing engine to render a plurality of image frames for left and right displays of a head mounted display (HMD); and foveation control hardware logic to independently control two or more of the plurality of graphics processing stages based on feedback received from an eye tracking module of the HMD, the feedback indicating a foveated region selected based on a current or anticipated direction of a user's gaze, the foveation control hardware logic to cause the two or more of the graphics processing stages to process the foveated region differently than other regions of the image frames.

Abstract: Systems and methods are provided to perform constrained BSDF sampling in relation to various algorithms, and specifically in relation to ray tracing algorithms. In some embodiments, a method is provided to generate samples by: determining a spherical polygon on a unit hemisphere; determining, on a unit circle, a projected area corresponding to the spherical polygon on the unit hemisphere; determining a parameterization of the projected area of the spherical polygon on the unit circle; generating samples in the projected area based on the parameterization; and generating samples in the spherical polygon. The unit circle is abase of the unit hemisphere, and the projection of the projected area is along a vector perpendicular to the unit circle. The generated samples in the spherical polygon correspond to the samples in the projected area. The method may further include evaluating a rendering equation based on the generated samples in the spherical polygon.

Abstract: Interactive systems using adaptive control software and hardware from known and later developed eyepieces to later developed head-wear to lenses, including implantable, temporarily insertable and contact and related film based types of lenses including thin film transparent elements for housing cameras lenses and projector and functional equivalent processing tools. Simple controls, real-time updates and instant feedback allow implicit optimization of a universal model while managing complexity.

Abstract: A fully-connected neural network may be configured for execution by a processor as a fully-fused neural network by limiting slow global memory accesses to reading and writing inputs to and outputs from the fully-connected neural network. The computational cost of fully-connected neural networks scale quadratically with its width, whereas its memory traffic scales linearly. Modern graphics processing units typically have much greater computational throughput compared with memory bandwidth, so that for narrow, fully-connected neural networks, the linear memory traffic is the bottleneck. The key to improving performance of the fully-connected neural network is to minimize traffic to slow “global” memory (off-chip memory and high-level caches) and to fully utilize fast on-chip memory (low-level caches, “shared” memory, and registers), which is achieved by the fully-fused approach.

Abstract: Methods and systems are disclosed for traversing nodes in a BVH tree by an intersection engine. Techniques disclosed comprise receiving, by the intersection engine, a traversal instruction, including a tracing-mode, ray data, and an identifier of a node to be traversed. Where the tracing-mode includes a closest hit mode and a first hit mode. If the node to be traversed is an internal node, the intersection engine determines, based on the tracing-mode, an order in which children nodes of the node are to be next traversed and output identifiers of the children nodes in the determined order.

Abstract: Apparatus and method for stack throttling.

Abstract: In one embodiment, a computing system may store, by first buffer blocks, texels organized into a texel array including a number of N×N texel sub-arrays. Each texel may fall within a corresponding N×N texel sub-array and may be associated with a two-dimensional sub-array coordinate indicating a position of that texel within the corresponding N×N texel sub-array. Each first buffer block of may be assigned a particular two-dimensional sub-array coordinate and stores a texel subset having the particular two-dimensional sub-array coordinate. The system may receive, by filter blocks, texels from the first buffer blocks. Each filter block may receive a texel from each first buffer block to form a corresponding N×N texel sub-array. The system may perform, by filter blocks, sampling operations parallelly on their respective N×N texel sub-arrays.