Abstract: Systems and methods of the present disclosure are directed to a method for optimizing utilization of graphics processors for machine learning inference tasks. The method includes simultaneously rendering, by a computing system comprising one or more computing devices, a plurality of textures from an input to a machine-learned model. The method includes generating, by the computing system, a plurality of shaders based at least in part on a layout of the plurality of textures, wherein each of the plurality of shaders corresponds to at least one operator of a plurality of operators of the machine-learned model. The method includes processing, by the computing system using a Graphics Processing Unit (GPU), the plurality of textures with the plurality of shaders to obtain a machine-learning output for the machine-learned model.

Abstract: Disclosed are an implant surgery planning method using automatic placement of an implant structure, a user interface providing method therefor, and a teeth image processing device therefor. One embodiment provides an implant surgery planning method using automatic placement of an implant structure, a user interface providing method therefor, and a teeth image processing device therefor, wherein during planning, placement of an implant structure is automatically performed with respect to all missing-tooth areas as well as a single missing-tooth area in a teeth image, and placement of an implant structure is automatically performed on the basis of information on relation with a surrounding structure of the missing-tooth area, whereby the present invention increases a user's convenience.

Abstract: A system and method for mapping location of one or more target data points to a D-dimensional environment are provided. The method includes mapping a sorted array of D-dimensional polygons to a D-dimensional cartesian coordinate system, determining a floor value and a ceiling value of each coordinate point of each target data point, calculating one or more shifted floor value and one or more shifted ceiling value of each coordinate point, combining the shifted floor value and the shifted ceiling value for determining one or more potential centroids, authenticating the potential centroids for determining one or more real centroids by determining the presence of the potential centroid in the array, calculating the distance of the target data point from the real centroids, and determining the polygon having the real centroid with the least distance from the target data point.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer-readable media that modify digital images via scene-based editing using image understanding facilitated by artificial intelligence. For example, in one or more embodiments the disclosed systems utilize generative machine learning models to create modified digital images portraying human subjects. In particular, the disclosed systems generate modified digital images by performing infill modifications to complete a digital image or human inpainting for portions of a digital image that portrays a human. Moreover, in some embodiments, the disclosed systems perform reposing of subjects portrayed within a digital image to generate modified digital images. In addition, the disclosed systems in some embodiments perform facial expression transfer and facial expression animations to generate modified digital images or animations.

Abstract: Methods, apparatus, devices, subsystems, and systems for holographically displaying live scenes including one or more three-dimensional (3D) objects are provided. In one aspect, a method includes capturing optical holograms of a live scene, digitizing/processing the optical holograms, and holographically reconstructing the live scene based on the digitized/processed holograms. In another aspect, a method includes capturing images/videos of a live scene, computing corresponding holograms, and holographically reconstructing the live scene based on the computed holograms.

Abstract: An optical display, including a first waveguide having a first set of surfaces, an input grating, a fold grating, and an output grating; an image input image node assembly; and a prismatic relay optics is provided. The prismatic relay optics may be configured to be optomechanically connected to the waveguide and the input image node assembly. The optical display is may also be configured to operate alone or as integrated with a headpiece to be used as a HUD. The HUD may have a first and a second configuration wherein the waveguide is decoupled or coupled.

Abstract: A three-dimensional (3D) mapping system can be configured to generate a 3D map of a real-world environment using annotation of large image data sets, in which terrestrial imagery can be programmatically labeled with accurate labels using remotely sensed overhead image data. The 3D mapping system can implement photogrammetry to create a point cloud. Each pixel in the point cloud can be classified based on a consensus of each frame. The point cloud can be co-registered to a remotely sensed reference dataset to provide precise spatial coordinates for each pixel. Different patches of point clouds can be stitched together to provide a complete 3D map for a given area, such as a downtown area of a city.

Abstract: Methods, systems and apparatuses may provide for technology that identifies first graphics data that is associated with spatially proximate positions. The technology identifies second graphics data that is associated with spatially proximate positions, and interleaves the first and the second graphics data across a plurality of storage tiles.

Abstract: A hierarchical acceleration structure for use in a ray tracing system. When generating a node for the hierarchical acceleration structure, the primitives in a particular portion of the 3D scene may be alternatively bounded by different shaped volumes. These bounding volumes or ‘bounding regions’ can be Axis Aligned Bounding Boxes (AABBs), although other bounding volumes can be used. The ray tracing system may use sets of two or more bounding volumes in a 3D scene to bound all the primitives within that portion. The choice of how to create sets of multiple bounding volumes within a portion of the 3D scene may be done by using a binary space partition (BSP). Different sets of bounding regions may present different amounts of surface area for a hypothetical ray entering the portion of the 3D scene dependent upon the expected ray direction or angle.

Abstract: A method renders photorealistic images in a web browser. The method is performed at a computing device having a general purpose processor and a graphics processing unit (GPU). The method includes obtaining an environment map and images of an input scene. The method also includes computing textures for the input scene including by encoding an acceleration structure of the input scene. The method further includes transmitting the textures to shaders executing on a GPU. The method includes generating samples of the input scene, by performing at least one path tracing algorithm on the GPU, according to the textures. The method also includes lighting or illuminating a sample of the input scene using the environment map, to obtain a lighted scene, and tone mapping the lighted scene. The method includes drawing output on a canvas, in the web browser, based on the tone-mapped scene to render the input scene.

Abstract: A compute-implemented method for determining an abnormal structure in an examination region in conjunction with an X-ray recording of an X-ray system, comprising receiving input data, the input data relating to an X-ray recording data set of the X-ray recording having multiple data channels; applying a trained function to the input data, the trained function being based on a machine learning method and applied to at least two data channels to determine the abnormal structure and generate output data; and providing the output data, the output data including an abnormal structure of the examination region.

Abstract: A coordinate generation system, a coordinate generation method, a computer readable recording medium with stored program, and a non-transitory computer program product are provided. The coordinate generation system includes processing units and a neural network module. The processing units are configured to obtain four vertex coordinates of an image. The vertex coordinates include first components and second components. The processing unit is configured to perform the following steps: obtaining first vector based on the first components of the four vertex coordinates and repeatedly concatenating the first vector so as to obtain a first input; obtaining second vector based on the second components of the four vertex coordinates and repeatedly concatenating the second vector so as to obtain a second input; and obtaining first output coordinate components and second output coordinate components of output coordinates based on the first input, the second input, and parameters of the neural network module.

Abstract: An electronic device and method for generation of reflectance maps for relightable 3D models is disclosed. The electronic device acquires multi-view image data that includes a set of images of an object and generates a 3D mesh of the object based on the multi-view image data. The electronic device obtains a set of motion-corrected images based on a minimization of a rigid motion associated with the object between images of the set of images and generates texture maps in a UV space based on the set of motion-corrected images and the 3D mesh. The electronic device obtains specular and diffuse reflectance maps based on a separation of specular and diffuse reflectance components from the texture maps, and obtains a relightable 3D model of the object based on the specular and diffuse reflectance maps.

Abstract: A background display system for a virtual image recording studio comprises a background display device which is configured to display, behind or above a real subject, a representation of a virtual background for a recording by means of an associated camera, and a control device which is configured to control the background display device. The control device comprises a data input for receiving lens data from the associated camera and is configured to adjust the representation of the virtual background in dependence of the received lens data.

Abstract: An electronic device is provided. The electronic device includes a signal transmission component, a first display portion and a second display portion. The signal transmission component extends along a first direction. The first display portion is electrically connected to the signal transmission component. The second display portion is electrically connected to the signal transmission component. The first display portion is separated from the signal transmission component by a first distance, the second display portion is separated from the signal transmission component by a second distance, and the first distance is less than the second distance.

Abstract: A back-end server acquires viewpoint information representing a position of a virtual viewpoint and a line-of-sight direction from the virtual viewpoint used to generate a virtual viewpoint image based on a plurality of images captured by a plurality of image capturing apparatuses, outputs a first virtual viewpoint image generated using three-dimensional geometric data representing the three-dimensional shape of an object for which the virtual viewpoint image is to be generated if the object is included in an area that is captured by the plurality of image capturing apparatuses, and outputs a second virtual viewpoint image generated without using the three-dimensional geometric data if the object is included in an area that is not captured by a subset of the plurality of image capturing apparatuses.

Abstract: A novel functional panel that is highly convenient or highly reliable is provided. The functional panel includes a first pixel. The first pixel includes a first element, a color conversion layer, and a first functional layer. The first functional layer is positioned between the first element and the color conversion layer. The first element has a function of emitting light and contains gallium nitride. The color conversion layer has a function of converting the color of light emitted from the first element into a different color. The first functional layer includes a first insulating film and a pixel circuit. The first insulating film includes a region positioned between the pixel circuit and the first element, and has an opening. The pixel circuit includes a first transistor. The first transistor includes a first oxide semiconductor film and is electrically connected to the first element through the opening.

Abstract: A method for outputting a grayscale data, a driving controller, and a display apparatus are provided. The method includes: in response to receiving a first input grayscale data which is not greater than a first threshold, generating a first output grayscale data without performing a dither computing on the first input grayscale data and outputting the first output grayscale data; in response to receiving a second input grayscale data which is greater than a second threshold, generating a first dithered grayscale data by performing a first dither computing on the second input grayscale data and outputting the first dithered grayscale data as a second output grayscale data. The second threshold is greater than the first threshold, and the second output grayscale data does not comprise any part of the second input grayscale data not processed by the first dither computing.

Abstract: The digital restaurant menu with touch screen comprises an electronic tablet, a plurality of operator controls, one or more wireless communication circuits, one or more speakers, and a microphone. The electronic tablet may be adapted for use in a restaurant and may present one or more menus to a user via a display screen and/or audibly via the one or more speakers. The electronic tablet may be adapted for use by a visually-impaired user. The one or more menus may be navigated using the plurality of operator controls, a touch sensitive overlay, voice commands spoken into the microphone, or combinations thereof. The plurality of operator controls may comprise braille relief. The electronic tablet may be operable to communicate with one or more external devices via the one or more wireless communication circuits.

Abstract: Techniques for a motion-based online interactive platform are described. The platform makes a motion-based online class realistic and allows a teacher to visualize motions performed by a student in a perspective and how close the motion is in view of an authoritative instructor (model). Each of computing devices used respectively by students is coupled to or includes a camera, where the camera is used by a student to show his/her presence or poses he/she performs. Data streams from the computing devices are received in a control computer associated with the teacher, where each of the data streams includes a video and a set of sensing data. A 3D avatar of a student is generated from the sensing data in the control computer and may be shown alone or along with an avatar of an instructor or model to visualize any differences between the student and the model in reference to a pose or motion.