Abstract: In various examples, to support training a deep neural network (DNN) to predict a dense representation of a 3D surface structure of interest, a training dataset is generated using a parametric mathematical modeling. A variety of synthetic 3D road surfaces may be generated by modeling a 3D road surface using varied parameters to simulate changes in road direction and lateral surface slope. In an example embodiment, a synthetic 3D road surface may be created by modeling a longitudinal 3D curve and expanding the longitudinal 3D curve to a 3D surface, and the resulting synthetic 3D surface may be sampled to form a synthetic ground truth projection image (e.g., a 2D height map). To generate corresponding input training data, a known pattern that represents which pixels may remain unobserved during 3D structure estimation may be generated and applied to a ground truth projection image to simulate a corresponding sparse projection image.

Abstract: An electronic apparatus performs a method of real time simulation of physical visual effect on one or more Graphics Processing Units (GPUs). The method includes a plurality of time steps. Each of the time steps includes: building up a mapping between particles and background grid blocks; sorting the particles to a level of granularity; transferring momenta and masses of the particles to grid nodes on the background grid blocks to compute forces on the grid nodes; updating velocities and resolving collisions from the computed forces on the grid nodes; and applying the updated velocities back to the particles from the grid nodes and advecting the particles. In some embodiments, the frequency of building up and sorting is reduced compared with the frequency of transferring, updating, and applying in the plurality of time steps.

Abstract: Systems and methods to reconfigure a scene based on timing and scene content are disclosed. Exemplary implementations may: obtain scene information defines a scene in which an event transpires; analyze the scene information that defines the event to identify effectual content and ineffectual content related to the event within the unedited event duration; and generate, based on the scene information and the analysis, a draft scene that includes an edited event comprising the effectual content from an edited event beginning to an edited event end in accordance with the prospective event duration.

Abstract: Devices, systems, and methods for augmented reality in surgical environments include an optical lens, which may be wearable by the user, and an augmented reality control system for generating an augmented reality presentation in the optical lens. The augmented reality presentation can depict the surgical area of the patient while artificially omitting one or more obstructions from the user's view point of the surgical area.

Abstract: An apparatus, such as a graphical processing unit (GPU), includes one or more processors configured to determine a plurality of first locality information of a received wave at a processing unit and to select a first processing element of a plurality of processing elements, the first processing unit having a plurality of second locality information from a previous wave that matches the plurality of first locality information to execute the received wave.

Abstract: An image display method includes drawing a first graph in a trusted execution environment (TEE). The method also includes displaying the first graph on a first layer in a display screen. The method further includes drawing a second graph in a rich execution environment (REE). The method additionally includes displaying the second graph on a second layer in the display screen. The first layer is located above the second layer. The first graph includes a transparent area and a non-transparent area. The non-transparent area includes content drawn in the TEE. A display area that is on the second graph and that corresponds to the transparent area includes content drawn in the REE. When the first layer and the second layer are overlapped for displaying, the transparent area does not shield the content drawn in the REE.

Abstract: Disclosed subject matter relates generally to graphics processing, and relates more particularly to processing graphics vertex content.

Abstract: An electronic device includes a first processing unit and a second processing unit, a first memory unit correspondingly set for the first processing unit and configured for data access by the first processing unit, and a second memory unit correspondingly set for the second processing unit and configured for data access by the second processing unit. The first processing unit occupies at least part of storage space of the second memory unit when a first criteria is met, and/or the second processing unit occupies at least part of storage space of the first memory unit when a second criteria is met.

Abstract: Systems, methods, and techniques dynamically utilize load balancing for workgroup assignments between a group of shader engines by a command processor of a graphics processing unit (GPU). Based on one or more commands received for execution, a plurality of workgroups is generated for assignment to a plurality of shader engines for processing, each shader engine including a respective quantity of active compute units. Each workgroup of the plurality of workgroups is dynamically assigned to a respective shader engine for execution based at least in part on indications of available resources respectively associated with each of the shader engines. In various embodiments, the indications of available resources may include physical parameters regarding each shader engine, as well as current status information regarding the processing of workgroups assigned to each shader engine.

Abstract: The present disclosure relates to an apparatus for controlling a multi-image display provided in a vehicle, and a method thereof. According to the present disclosure, a multi-image display device may simultaneously output two or more images separated depending on an output angle or an optical frequency, and a first reflection member provided in an indoor ceiling of the vehicle may reflect one image among the two or more images output from the multi-image display device.

Abstract: There is described a method of shading a group of pixels in a fragment shader in a raster graphics pipeline. At least one first pilot pixel of the group of pixels is shaded under a first precision. At least one second pilot pixel of the group of pixels is shaded under a second precision. An error value representing a difference between the first and second pilot pixels is calculated. At least one other pixel of the group of pixels is shaded under the first precision if the error value is greater than an error threshold. The at least one other pixel is shaded under the second precision if the error value is smaller than the error threshold.

Abstract: This disclosure provides systems, devices, apparatus, and methods, including computer programs encoded on storage media, for GPU wave-to-wave optimization. A graphics processor may execute a shader program for a first wave associated with a draw call or a compute kernel. The graphics processor may identify at least one first indication for the first wave associated with the draw call or the compute kernel. The graphics processor may store the at least one first indication for the first wave to a memory location. The graphics processor may execute the shader program for at least one second wave associated with the draw call or the compute kernel. The execution of the shader program for the at least one second wave may be based on the shader program for the at least one second wave reading the memory location to retrieve the at least one first indication.

Abstract: Embodiments of this disclosure include an information processing method, an information processing apparatus, and non-transitory computer-readable storage medium. A first key frame that includes initial posture data corresponding to an initial posture of a target virtual character is obtained. Target posture data is determined by inputting the initial posture data and a target task to a policy network trained by reinforcement learning. An output of the policy network indicates the target posture data corresponding to a target posture. At least one force to be acted on at least one first joint of the target virtual character is determined according to the initial and the target posture data. A posture of the target virtual character is adjusted from the initial posture to the target posture by applying the at least one force on the at least one first joint to obtain a second key frame.

Abstract: A human face data processing method according to an embodiment of the present disclosure includes acquiring a picture of a human face by means of a scanning apparatus, obtaining point cloud information by means of a structured light stripe, and further obtaining a three-dimensional model of the human face, and mapping the three-dimensional model onto a circular plane in an area-preserving manner so as to form a two-dimensional human face image. Three-dimensional data is converted into two-dimensional data, thereby facilitating data storage. In addition, the three-dimensional data uses the area-preserving manner, such that the restoration quality is better when the two-dimensional data is restored to the three-dimension data, thereby facilitating the re-utilization of a three-dimensional image.

Abstract: Embodiments provide mechanisms to facilitate compute operations for deep neural networks. One embodiment comprises a graphics processing unit comprising one or more multiprocessors, at least one of the one or more multiprocessors including a register file to store a plurality of different types of operands and a plurality of processing cores. The plurality of processing cores includes a first set of processing cores of a first type and a second set of processing cores of a second type. The first set of processing cores are associated with a first memory channel and the second set of processing cores are associated with a second memory channel.

Abstract: In various examples, to support training a deep neural network (DNN) to predict a dense representation of a 3D surface structure of interest, a training dataset is generated using a parametric mathematical modeling. A variety of synthetic 3D road surfaces may be generated by modeling a 3D road surface using varied parameters to simulate changes in road direction and lateral surface slope. In an example embodiment, a synthetic 3D road surface may be created by modeling a longitudinal 3D curve and expanding the longitudinal 3D curve to a 3D surface, and the resulting synthetic 3D surface may be sampled to form a synthetic ground truth projection image (e.g., a 2D height map). To generate corresponding input training data, a known pattern that represents which pixels may remain unobserved during 3D structure estimation may be generated and applied to a ground truth projection image to simulate a corresponding sparse projection image.

Abstract: A tag buffer implements a tag buffer stage of a rasterization phase in a tile-based rendering graphics processing system having a rendering space sub-divided into a plurality of tiles to which primitives can be associated. A buffer stores an identifier that identifies a visible primitive fragment at each sample position of a tile of the plurality of tiles. A look-up table stores an entry for transformed primitive blocks that indicates whether the tag buffer has received information identifying a primitive fragment associated with that transformed primitive block.

Abstract: The present invention relates to a method, a computer program with instructions, and a device for generating a display of an augmented reality head-up display for a motor vehicle. In a first step, a position of the motor vehicle is determined. A plurality of virtual elements for a navigation hint is then provided on the basis of the position of the motor vehicle. The virtual elements, as seen from the motor vehicle, extend on both sides in a lateral direction to outside a field of view of the augmented reality head-up display. Finally, at least that virtual element which has the shortest distance to a center of the field of view is rendered.

Abstract: Systems and methods for doing presenting an avatar that speaks sign language based on sentiment of a speaker is disclosed herein. A translation application running on a device receives a content item comprising a video and an audio, wherein the audio comprises a first plurality of spoken words in a first language. The video comprises a character speaking the first plurality of spoken words in the first language. The translation application translates the first plurality of spoken words of the first language into a first sign of a first sign language. The translation application determines an emotional state expressed by the character based on sentiment analysis. The translation application generates an avatar that speaks the first sign of the first sign language where the avatar exhibits the determined emotional state. The content item and the avatar are presented for display on the device.

Abstract: A multicore graphics processing unit (GPU) and a method of operating a GPU are provided. The GPU comprises at least a first core and a second core. At least one of the cores in the multicore GPU comprises a master unit configured to distribute geometry processing tasks between at least the first core and the second core.