Abstract: One embodiment provides for a graphics processing unit comprising a processing cluster to perform multi-rate shading via coarse pixel shading and output shaded coarse pixels for processing by a post-shader pixel processing pipeline.

Abstract: Embodiments of the present disclosure provide a video processing method, a video processing apparatus, an electronic device and a computer-readable storage medium. The video processing method includes: displaying an initial image which includes a first-style image; in response to a first trigger event, displaying an image switching animation which is used for presenting a dynamic process of the switching from the initial image to a target image which includes a second-style image; and in response to completion of the displaying of the image switching animation, displaying the target image. A switching image in the image switching animation includes a first image area, a second image area and a third image area, and the first image area covers the entire image area of the image switching animation by means of position movement and in a time-sharing manner, and has a change in shape during a position movement process.

Abstract: Systems and methods for retrieval of augmented parameters for an artificial intelligence (AI)-based character are provided. An example method includes receiving, from a user via a user interface, at least one keyword describing the AI-based character; retrieving, from at least one data source and based on the at least one keyword, the augmented parameters describing the AI-based character; and generating, based on the augmented parameters, an AI-based character model corresponding to the AI-based character. The at least one data source includes a database configured to store records associated with the AI-based character, an online search service, and a set of clusters associated with a type of a feature of the AI-based character and at least one hidden prompt corresponding to the type of the feature. The type of the feature includes one of the following: a voice, a dialog style, an emotional state, an age, and temperament.

Abstract: Methods and systems are described. A system includes a redundant shader pipe array that performs rendering calculations on data provided thereto and a shader pipe array that includes a plurality of shader pipes, each of which performs rendering calculations on data provided thereto. The system also includes a circuit that identifies a defective shader pipe of the plurality of shader pipes in the shader pipe array. In response to identifying the defective shader pipe, the circuit generates a signal. The system also includes a redundant shader switch. The redundant shader switch receives the generated signal, and, in response to receiving the generated signal, transfers the data for the defective shader pipe to the redundant shader pipe array.

Abstract: Disclosed is a system for encoding and/or rendering animations without temporal or spatial restrictions. The system may encode an animation as a point cloud with first data points having a first time value and different positional and non-positional values, and second data points having a second time value and different positional and non-positional values. Rendering the animation may include generating and presenting a first image for the first time value of the animation based on the positional and non-positional values of the first data points, and generating and presenting a second image for the second time value of the animation by changing a visualization at a first position in the first image based on the positional values of a data point from the second data points corresponding to the first position and the data point non-positional values differing from the visualization.

Abstract: A computer-implemented method that allows users evaluate the densities of images and search for abnormalities in three-dimensional space. The voxel buildup uses a series of two-dimensional images and evaluates every pixel based on the user described predetermined threshold value at runtime. A singular optimized voxel-generated mesh is spawned to represent the combined locations of every pixel.

Abstract: Technologies for 3D virtual environment placement of 3D models based on 2D images are disclosed. At least an outline of a 3D virtual environment may be generated. A 2D image of one or more 2D images may be identified. A first product from the first 2D image may be identified. At least one 3D model of one or more 3D models based, at least, on the first product may be determined. A first location for placement of the first product in the 3D virtual environment may be identified. The at least one 3D model may be added within the 3D virtual environment based, at least, on the first location. The 3D virtual environment may be rendered into a visually interpretable form. A second product may be identified from the first 2D image, forming a first grouping of products. A starting element for the first grouping of products may be determined.

Abstract: Systems and methods disclosed provided a virtual reality experience, including: a set of motorized pads for coupling to feet of a user; a means of communication between the set of motorized pads and a computing environment operating a virtual reality headset; such that the set of motorized pads are configured to provide pressure on a user's feet as an avatar within a virtual environment traverses the environment. Systems and methods disclosed further provide a multiuser virtual reality experience. Systems and methods disclosed further provide a multiuser experience including generation of a common environment viewable in the virtual reality headsets; and recording a tour including audio data as the first user describes the common environment.

Abstract: Embodiments of the present disclosure provide a method for audio-driven character lip sync, a model for audio-driven character lip sync, and a training method therefor. A target dynamic image is obtained by acquiring a character image of a target character and speech for generating a target dynamic image, processing the character image and the speech as image-audio data that may be trained, respectively, and mixing the image-audio data with auxiliary data for training. When a large amount of sample data needs to be obtained for training in different scenarios, a video when another character speaks is used as an auxiliary video for processing, so as to obtain the auxiliary data. The auxiliary data, which replaces non-general sample data, and other data are input into a model in a preset ratio for training. The auxiliary data may improve a process of training a synthetic lip sync action of the model, so that there are no parts unrelated to the synthetic lip sync action during the training process.

Abstract: A display method, a displaying device, electronic equipment and a storage medium. The display method comprises: acquiring menu data, target display position data of the menu data on a display screen, and source data, wherein the source data includes N channels of first display data, and the first display data has a preset size; conducting format conversion on the menu data to obtain M channels of second display data with a preset size; fusing the first display data with the second display data according to the target display position to obtain third display data; and displaying the third display data on the display screen.

Abstract: The technology described in this document can be embodied in a method of displaying images of portions of a human body on a display device. The method includes receiving a representation of a plurality of images that includes images of at least two different modalities, and location information corresponding to at least a subset of the plurality of images. A first image of a first modality is displayed on the display device in accordance with the corresponding location information. A second image of a second modality is overlaid on the first image in accordance with corresponding location information. At least a third image is overlaid on the first image in accordance with corresponding location information, the third image being of the second modality, and the second and third images being displayed concurrently for at least a period of time.

Abstract: Systems and methods to improve machine learning by explicitly over-fitting environmental data obtained by an imaging system, such as a monocular camera are disclosed. The system includes training self-supervised depth and pose networks in monocular visual data collected from a certain area over multiple passes. Pose and depth networks may be trained by extracting data from multiple images of a single environment or trajectory, allowing the system to overfit the image data.

Abstract: Devices, systems and processes for an integrated internal and external camera system that enhances the passenger experience in vehicles are described. One example method for enhancing a passenger experiences includes capturing a first set of images of an area around the vehicle using an external camera system, capturing a second set of images of one or more passengers inside the vehicle using an internal camera system, recognizing at least one gesture made by the one or more passengers based on the second set of images, identifying an object or a location external to the vehicle based on the first set of images and the at least one gesture, and displaying information related to the object or the location to the one or more passengers.

Abstract: In an exemplary process for specifying an entrance or exit effect in a computer-generated reality environment, in response to a user entering or exiting the computer-generated reality environment, a transition effect is provided.

Abstract: Systems and methods disclosed provided a virtual reality experience, including: a set of motorized pads for coupling to feet of a user; a means of communication between the set of motorized pads and a computing environment operating a virtual reality headset; such that the set of motorized pads are configured to provide pressure on a user's feet as an avatar within a virtual environment traverses the environment. Systems and methods disclosed further provide a multiuser virtual reality experience. Systems and methods disclosed further provide a multiuser experience including generation of a common environment viewable in the virtual reality headsets; and recording a tour including audio data as the first user describes the common environment.

Abstract: An information processing device including a display unit, a detector, and a first control unit and a method of using same. The display unit may be a head-mounted display. The display unit is capable of providing the user with a field of view of a real space and a virtual object. The detector detects an azimuth of the display unit around at least one axis and display of the virtual object is controlled based in the detected azimuth.

Abstract: A point cloud data processing method and device, a computer device and a storage medium are provided. The method includes: acquiring point cloud data, and constructing a corresponding neighboring point set for each of data points in the point cloud data; calculating Hausdorff distances between the neighboring point set and a pre-constructed kernel point cloud to obtain a distance matrix; calculating a convolution of the neighboring point set with the distance matrix and a network weight matrix in a Hausdorff convolution layer in an encoder, to obtain high-dimensional point cloud features, the encoder and a decoder being two parts in a deep learning network; and reducing feature dimension of the high-dimensional point cloud features through the decoder, so that a classifier performs semantic classification on the point cloud data according to object point cloud features obtained by the dimension reduction.

Abstract: A computing system captures markerless motion data of a user via a camera of the computing system. The computing system retargets the first plurality of points and the second plurality of points to a three-dimensional model of an avatar associated with the user, wherein the avatar is associated with an identity non-fungible token that uniquely represents the user across Web2 environments and Web3 environments, and wherein retargeting the first plurality of points and the second plurality of points animates the three-dimensional model of the avatar. The computing system renders a video local to the computing system, wherein the video comprises the markerless motion data of the user retargeted to the three-dimensional model of the avatar causing hands, face, and body of the avatar to be animated in real-time. The computing system causes a non-fungible token to be generated, the non-fungible token uniquely identifying ownership of the video.

Abstract: Provided are a material presentation method and apparatus, a terminal and a storage medium. The material presentation method includes steps described below. A to-be-presented splash presentation material is received. A splash presentation image of the to-be-presented splash presentation material is magnified, and the magnified splash presentation image is presented. In response to the presented magnified splash presentation image satisfying a minifying condition, the presented magnified splash presentation image is minified.

Abstract: Implementations described herein relate to methods, systems, and computer-readable media for remote production collaboration tools. The remote production collaboration tools can include one or more client devices, a server providing a single connection point, and an animation server configured to produce an animated production. The one or more client devices may provide motion capture data, audio data, control data, and/or associated timestamps. The animation server is configured to process the motion capture data, audio data, control data, and/or the associated timestamps and create the animated production. The animated production may be transmitted as a video stream.