Abstract: The method automates the creation of explainer videos from text-based data by using a processor to analyse and identify components of an input document, such as text and images. The processor determines the reading order based on visual analysis and extracts structured information, including text-based and image-based elements. The text-based elements include paragraphs, lines, tables, complex formulas and the like. The image-based elements include schematics, graphs, illustrations and the like. The processor identifies the optimal amount of content displayable per screen, and the associated timing selects a suitable layout from predefined layouts. It generates an explainer video for at least one topic in the input document, ensuring a coherent and visually organised presentation.

Abstract: A processor-implemented method included determining first color values of a portion of pixels of plural pixels, the portion of pixels being in a tile frame corresponding to a partial region of an input frame, by performing shading using a shader module on the portion of pixels included, determining second color values of other pixels, the other pixels being pixels of the plural pixels not included in the portion of pixels of the tile frame, by performing neural network-based super-sampling processing on the other pixels, and determining a rendered tile frame including the first color values of the portion of pixels and the second color values of the other pixels, the determining of the first color values of the portion of pixels including determining edge color values of pixels in an edge region by performing shading using the shader module on the pixels in the edge region of the tile frame.

Abstract: It is desirable to provide a technology capable of more accurately estimating a motion of a three-dimensional model between a plurality of frames. Provided is an information processing apparatus including a movement amount calculation unit that calculates a movement amount associated with a first vertex included in a first frame on the basis of statistical processing according to color information associated with the first vertex, color information associated with a second vertex included in a second frame after the first frame, three-dimensional coordinates associated with the first vertex, and three-dimensional coordinates associated with the second vertex.

Abstract: A video processing method and apparatus, a device, and a storage medium are provided. The method includes: acquiring three-dimensional reconstruction data of a video image and a first texture image corresponding to a current moment; determining target brightness of a target pixel position on the first texture image corresponding to the current moment according to a pre-obtained mapping relationship between pixel positions and initial brightness and brightness change rates; adjusting brightness of the target pixel position on the first texture image to the target brightness so as to obtain a second texture image; and mapping the second texture image onto the video image based on the three-dimensional reconstruction data so as to obtain a target video image.

Abstract: In an approach to improve computer-based virtual world collaboration environments, embodiments of the present invention identifies, by a client computer, a structure of a virtual world collaboration room and the placement of participants in the virtual world collaboration room, and correlates, by an internet of things (IoT) sensor set, body language of an avatar to match a spoken context of the avatar. Further, embodiments select personalized virtual world collaboration room or a predetermined physical location to conduct a virtual world collaboration and utilize a generative adversarial network (GAN) to adapt the avatar to the personalized virtual world collaborative environment. Additionally, embodiments perform real-time adaptation, by the GAN, of participating avatars to generate and output a required body language for the participating avatars based on their location different personalized virtual world collaboration rooms.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for automated image generation. Methods can include generating, by an artificial intelligence (“AI”) system, image prompts based on (i) a category and (ii) a set of constraints. At least one of the image prompts is removed based on a set of excluded terms that are not allowed to be included in qualified image prompts. A set of images is generated using the set of qualified image prompts. One or more images that include excluded objects are removed to obtain a set of qualified generated images. The set of qualified generated images are stored with data specifying the category. At least one qualified generated image is combined with other content, resulting in a new digital component. The new digital component is transmitted to a client device in response to a request for content in the category.

Abstract: Provided are a method and device for processing an oral image according to embodiments. The oral image processing method includes displaying adjacent tooth data adjacent to preparation tooth data and restoration data for restoration of the preparation tooth data, receiving an input for adjusting adaptation between the restoration data and the adjacent tooth data, processing the adaptation between the restoration data and the adjacent tooth data in response to the received input, and displaying the restoration data adapted according to the processed adaptation.

Abstract: A method of generating input data for a machine learning model includes determining, for a sensor, a point cloud with points detected by the sensor from surfaces in the environment of the sensor, generating a preliminary target sensor point cloud for a target sensor by transforming, for the sensor, points of the determined point cloud into points from the perspective of the target sensor according to the relative position of the target sensor to the sensor, generating a target sensor point cloud for the target sensor by using the preliminary target sensor point cloud, wherein points which, due to one or more surfaces for which points exist in the preliminary target sensor point cloud, are not detectable by the target sensor are eliminated in the target sensor point cloud, and using the target sensor point cloud as input for the machine learning model.

Abstract: Techniques include receiving a first image frame showing an object from a first view and a second image frame showing the object from the first view, wherein the first image frame and second image frame are different image frames. The techniques further include generating, based at least in part on the first image frame, a third image frame of the object from a second view that is different than the first view. The techniques further include generating, based at least in part on the second image frame and the third image frame, a fourth image frame of the object from the second view. The techniques further include generating a video of the object from the second view by combining at least the third image frame and fourth image frame.

Abstract: In various examples, systems and methods are disclosed that relate to data augmentation for training/updating perception models in autonomous or semi-autonomous systems and applications. For example, a system may receive data associated with a set of frames that are captured using a plurality of cameras positioned in fixed relation relative to the machine; generate a panoramic view based at least on the set of frames; provide data associated with the panoramic view to a model to cause the model to generate a high dynamic range (HDR) panoramic view; determine lighting information associated with a light distribution map based at least on the HDR panoramic view; determine a virtual scene; and render an asset and a shadow on at least one of the frames, based at least on the virtual scene and the light distribution map, the shadow being a shadow corresponding to the asset.

Abstract: In an embodiment, operations include receiving a graph dataset including a set of graphs. The operations further include generating, by a task scheduler, a set of task queues configured to process the received graph dataset in parallel. The operations further include determining, by the set of task queues, a set of graph attributes for each graph of the set of graphs. The operations further include receiving, by a web service, one or more graph attributes of the determined set of graph attributes. The operations further include transmitting, by the web service, attribute information including the received one or more graph attributes to a client browser. The client browser is configured to determine a set of graph views based on the transmitted attribute information. The operations further include controlling rendering of the determined set of graph views on the client browser.

Abstract: An interface display method and an electronic device are provided. The method includes: performing screen-on interception in response to a screen wakeup operation performed by a user on an electronic device, to prevent a screen from being lit up; drawing an interface adapted to a screen orientation; and after the interface is drawn, disabling screen-on interception, and displaying the interface based on the screen orientation. The electronic device may collect and report data in real time through a sensor, and calculate a current screen orientation in real time, so that the drawn interface adapts to the current screen orientation.

Abstract: An avatar generation system includes an acquisition unit, an execution unit, a detection unit, and an output unit. The acquisition unit acquires setting data for bones and meshes of an avatar. The execution unit executes processing of changing a posture of the avatar based on the setting data and a defined movement. The detection unit detects a state of penetration between the meshes constituting the avatar based on data in which the posture of the avatar has been changed. The output unit outputs a result of the detection.

Abstract: An improved algorithm and technique for use in interactive graphics is disclosed. The new algorithm works by keeping three values in fast, memoryless pixel local storage for coverage counting (store the current coverage count at the pixel being covered), original framebuffer color (stores the color that was in the framebuffer at the pixel being covered immediately before the current path starts rendering), and the path ID (stores the unique ID of the last path to be drawn at the current pixel). The framebuffer is accessed from a fourth, texture-backed, pixel local storage plane.

Abstract: An image display method and apparatus are provided. The method includes: determining a target display position of a target object in an original image and a first size of a display region of an ink screen (102); generating a monochrome image with the first size according to the original image, where color values of pixel points in the monochrome image include at least one standard color value, and the target object is located at a target display position in the monochrome image (104); sending the monochrome image to the ink screen, so that the monochrome image is displayed by the ink screen in the display region (106). This method can improve the definition of the monochrome image displayed by the ink screen.

Abstract: Generating a three-dimensional virtual representation of a three-dimensional physical object can be based on capturing or receiving a capture bundle or a set of images. In some examples, generating the virtual representation of the physical object can be facilitated by user interfaces for identifying a physical object and capturing a set of images of the physical object. Generating the virtual representation can include previewing or modifying a set of images. In some examples, generating the virtual representation of the physical object can include generating a first representation of the physical object (e.g., a point cloud) and/or generating a second three-dimensional virtual representation of the physical object (e.g., a mesh reconstruction). In some examples, a visual indication of the progress of the image capture process and/or the generation of the virtual representation of the three-dimensional object can be displayed, such as in a capture user interface.

Abstract: Distributed real-time rendering is a cutting-edge solution to address the ever-increasing demands for realistic and immersive visual experiences, such as provided by augmented or artificial reality applications. Embodiments of the present specification provide a method and a system for distributed real-time rendering of a scene. The rendering is distributed among a root instance and a plurality of node instances. The root instance receives the scene to render, segments it into a plurality of objects, computes lighting parameters, assigns rendering tasks associated with objects to node instances, and communicates object metadata and the lighting parameters to the assigned node instance. The node instances render objects and create point clouds and u, v maps, which are then used by the root instance to render the scene for real-time display. The method and system can be used by end devices that include position sensors, processors, communication means and a display to provide immersive experiences.

Abstract: Disclosed is a circuit and a method for video data conversion and a display device. The circuit comprises: a pixel splicing module, to sequentially receive each group of pixel data in each frame of image data in input video data, to perform pixel splicing on the received pixel data, and to output spliced pixel data; a data conversion module, to perform data processing on the spliced pixel data to convert each frame of image data into first image data representing a first part of the corresponding frame of image and second image data representing a second part of the corresponding frame of image. The first part at least comprises a left half part of the frame of image, the second part at least comprises a right half part of the frame of image. Data throughput requirements can be met while dynamic contrast and sharpening processing is performed.

Abstract: In one embodiment, a method includes accessing, by a computing device, a natural-language input comprising a description of an avatar and generating, from the natural-language input and by a trained avatar-creation model, a set of avatar-attribute feature vectors. The method further includes determining, by a trained avatar-attribute classifier, one or more avatar attributes from the set of avatar-attribute feature vectors; and generating the avatar based on the determined one or more avatar attributes for presentation on a display of a computing device.

Abstract: A method for generating a set of adversarial patches for an image. The method includes segmenting the image into a plurality of regions; selecting a set of target regions that satisfies an attacking criterion by discretely searching of the plurality of regions; and generating a set of adversarial patches by using the set of target regions.