Abstract: Generative shape creation and editing is leveraged in a digital medium environment. An object editor system represents a set of training shapes as sets of visual elements known as “handles,” and converts sets of handles into signed distance field (SDF) representations. A handle processor model is then trained using the SDF representations to enable the handle processor model to generate new shapes that reflect salient visual features of the training shapes. The trained handle processor model, for instance, generates new sets of handles based on salient visual features learned from the training handle set. Thus, utilizing the described techniques, accurate characterizations of a set of shapes can be learned and used to generate new shapes. Further, generated shapes can be edited and transformed in different ways.

Abstract: Disclosed is a system and associated methods for generating a composite image from scans or images that are aligned using invisible fiducials. The invisible fiducial is a transparent substance or a projected specific wavelength that is applied to and changes reflectivity of a surface at the specific wavelength without interfering with a capture of positions or visible color characteristics across the surface. The system performs first and second capture of a scene with the surface, and detects a position of the invisible fiducial in each capture based on values measured across the specific wavelength that satisfy a threshold associated with the invisible fiducial. The system aligns the first capture with the second capture based on the detected positions of the invisible fiducial, and generates a composite image by merging or combining the positions or visible color characteristics from the aligned captures.

Abstract: The present invention discloses a method and system for directed transfer of cross-domain data based on high-resolution remote sensing images. In the method of the present invention, first, an objective loss function which combines an image translation loss and a model adaptive loss of an image translation network model is established, thus overcoming the technical shortcoming that an existing data translation technique fails to take a specific task into full consideration and ignores a negative impact of data translation on the specific task. Further, a trained image translation network model is fine-tuned based on sample data, so that the image translation network model keeps translation towards the effect desired by the target model, thus avoiding over-interpretation or over-simplification during directed transfer of cross-domain data and improving accuracy of directed transfer of the cross-domain data based on the high-resolution remote sensing images.

Abstract: A colour processor for mapping an image from source to destination colour gamuts has an input for receiving a source image including a plurality of source colour points expressed according to the source gamut; a colour characterizer configured to, for each source colour point in the source image, determine a position of intersection of a curve with the boundary of the destination gamut; and a gamut mapper configured to, for each source colour point in the source image: if the source colour point lies inside the destination gamut, apply a first translation factor to translate the source colour point to a destination colour point within a first range of values; or if the source colour point lies outside the destination gamut, apply a second translation factor, different to the first translation factor, to translate the source colour point to a destination colour point within a second range of values.

Abstract: A method of deferred vertex attribute shading includes computing, at a graphics processing pipeline of a graphics processing unit (GPU), a plurality of vertex attributes for vertices of each primitive of a set of primitives. The plurality of vertex attributes to be computed includes a vertex position attribute and at least a first non-position attribute for each primitive. One or more primitives of the set of primitives that do not contribute to a rendered image are discarded based upon the vertex position attribute for vertices of the set of primitives. A set of surviving primitives is generated based on the culling and deferred attribute shading is performed for at least a second non-position attribute for vertices of the set of surviving primitives.

Abstract: The system for recognizing one or more objects of a design creative within an environment, analyzing the one or more objects using a deep neural networking model and generating a modified design creative by (i) determining a location of a design creative within the media content, (iii) determining an object from the design creative, (iv) determining an attribute of the object, (v) implementing a compliance rule to the attribute of the object to determine a distinctness and an effectiveness of a brand product, (vi) generating an attention sequence and heatmap for the media content, (vii) automatically generating a first recommendation based on the compliance rule, the attention heatmap, and the attention sequence, and (viii) automatically generating a modified design creative for the environment based on the attention heatmap, the attention sequence and the generated first recommendation using the deep neural networking model.

Abstract: Disclosed herein are a slim immersive display device and a slim visualization device. The slim immersive display device includes a slim visualization module for forming an image on a retina of an eyeball of a user based on an externally input image signal, a state information acquisition unit for acquiring a state of an external device as a state image, and a content control unit for analyzing the state image, generating an image corresponding to virtual-reality environment information, and inputting the image to the slim visualization module.

Abstract: An encoder is disclosed that uses hyperspectral data to produce a unified three-dimensional (“3D”) scan that incorporates depth for various points, surfaces, and features within a scene. The encoder may scan a particular point of the scene using frequencies from different electromagnetic spectrum bands, may determine spectral properties of the particular point based on returns measured across a first set of bands, may measure a distance of the particular point using frequencies of another band that does not interfere with the spectral properties at each of the first set of bands, and may encode the spectral properties and the distance of the particular point in a single hyperspectral dataset. The spectral signature encoded within the dataset may be used to classify the particular point or generate a point cloud or other visualization that accurately represents the spectral properties and distances of the scanned points.

Abstract: A computer-implemented method and system for virtual visualization of overall regions are disclosed. Images of a overall region may be obtained by a server and used to generate a virtual model of the overall region. The server may generate a virtual environment that includes the virtual model of the overall region. A user may indicate a specific area of the virtual environment. The server may then dispatch an imaging vehicle to the location of the indicated area to capture additional image data representative of the indicated area. This additional set of image data may be used by the server to generate higher resolution virtual models of the indicated area that are integrated into the virtual environment.

Abstract: One embodiment provides for a graphics processor comprising a block of graphics compute units, a graphics processor pipeline coupled to the block of graphics compute units, and a programmable neural network unit including one or more neural network hardware blocks. The programmable neural network unit is coupled with the block of graphics compute units and the graphics processor pipeline. The one or more neural network hardware blocks include hardware to perform neural network operations and activation operations for a layer of a neural network. The programmable neural network unit can configure settings of one or more hardware blocks within the graphics processor pipeline based on a machine learning model trained to optimize performance of a set of workloads.

Abstract: Aspects of the technology described herein relate to techniques for guiding an operator to use an ultrasound device. Thereby, operators with little or no experience operating ultrasound devices may capture medically relevant ultrasound images and/or interpret the contents of the obtained ultrasound images. For example, some of the techniques disclosed herein may be used to identify a particular anatomical view of a subject to image with an ultrasound device, guide an operator of the ultrasound device to capture an ultrasound image of the subject that contains the particular anatomical view, and/or analyze the captured ultrasound image to identify medical information about the subject.

Abstract: Provided is an eyewear display system includes: a scanner including a measuring unit configured to acquire three-dimensional coordinates, a point cloud data acquiring unit configured to acquire point cloud data, and a communication unit; an eyewear device including a display, a relative position detection sensor, a relative direction detection sensor, and a communication unit; a storage device including CAD design data of an observation site; and a data processing device configured to synchronize a coordinate space of the scanner, a coordinate space of the eyewear device, and a coordinate space of the CAD design data, and convert observation data OD and/or observation data prediction PD into data in the coordinate space of the CAD design data, such that the eyewear device displays the observation data OD or observation data prediction PD on the display by superimposing the observation data OD or observation data prediction PD on an actual landscape.

Abstract: Systems and methods for managing user input for a remote application session between a client device and a remote server are provided. Input command data is received from a client device for an application executing within a remote session established between the client device and the remote server. The input command data specifies a plurality of points within a graphical user interface where the input command data was captured on the client device. Spline curves are calculated based on a first subset of sequential points from the plurality of points. Simulated input events are generated that specify intermediate points calculated along the spline curves. The simulated input events are injected into the remote server for the application.

Abstract: Implementations are described herein for automatically generating synthetic training images that are usable, for instance, as training data for training machine learning models to detect and/or classify various types of plant diseases at various stages in digital images. In various implementations, one or more environmental features associated with an agricultural area may be retrieved. One or more synthetic plant models may be generated to visually simulate one or more stages of a progressive plant disease, taking into account the one or more environmental features associated with the agricultural area. The one or more synthetic plant models may be graphically incorporated into a synthetic training image that depicts the agricultural area.

Abstract: Systems and methods are disclosed for generating a 3D avatar using a biomechanical analysis of observed actions with a focus on representing actions through computer-generated 3D avatars. Physical quantities of biomechanical actions can be measured from the observations, and the system can analyze these values, compare them to target or optimal values, and use the observations and known biomechanical capabilities to generate 3D avatars.

Abstract: A method includes displaying, on a display, a representation of a physical environment and a computer-generated object. The method includes generating a three-dimensional (3D) point cloud associated with the representation of the physical environment. The method includes obtaining depth information characterizing the physical environment. The method includes obtaining a capture event associated with a capture region within the representation of the physical environment. The capture region includes a portion of the computer-generated object. The method includes, in response to obtaining the capture event, disambiguating a group of points from the 3D point cloud, and generating, based on a function of the depth information and the group of points, a volumetric representation of the capture region. The group of points satisfies a spatial threshold with respect to the capture region. The volumetric representation includes a volumetric representation of the portion of the computer-generated object.

Abstract: Computer-implemented methods and systems for curating virtual models and populating overlays within a virtual environment are described herein. A server may receive a data request from a user electronic device. The data request may comprise a property of interest located at a particular portion of an overall region. The server may then dynamically acquire a virtual model for rendering the property within a virtual environment at the user electronic device based on the data request. The server may then curate the virtual model in accordance with rules that emphasize features associated with the property that are relevant to assessing risks associated with the property when assessing the property. The server may then identify the curated property modeled by the virtual model, obtain annotation records associated with the features of the property, and populate an annotations overlay rendered in the virtual environment with information included in the annotation records.

Abstract: Three-dimensional occlusion can be used when generating AR display overlays. Depth information can be used to delete portions of an AR element, based on intervening objects between a viewer and the AR element. In cases where the depth information does not impart a complete picture of the intervening objects, additional image processing and object detection systems and techniques can be used to further improve the precision of the occlusion.

Abstract: A system for generating missing data for 3D models of intraoral structures of a patient. The system may include a hand-held intraoral scanner and a computer having instructions that the system to scan the intraoral structure of a patient to generate first 3D data of the surface of a prepared tooth, generate a 3D virtual model of the prepared tooth based on the first 3D data, determine the 3D virtual model is missing a surface associated with the prepared, generate second 3D data approximating the surface that is determined to be missing from the 3D virtual model intraoral structure that is missing, and combine the second 3D data with the 3D virtual model such that the 3D virtual model includes the approximated surface, wherein the approximated surface approximates the prepared tooth associated with the surface determined to be missing from the 3D virtual model.

Abstract: A server includes processing circuitry configured to receive one or more images, the one or more images including one or more representations of people. Additionally, the processing circuitry is configured to apply a neural network to the one or more images, wherein the neural network classifies at least one aesthetic component of each image of the one or more images, an aesthetic component score being generated for each image in the one or more images. Further, the processing circuitry is configured to generate a user eyewear equipment profile for a user, the user being matched to a persona from a personae database, each persona in the personae database being linked to one or more persona eyewear equipment profiles, the one or more persona eyewear equipment profiles being based on the aesthetic component score, and select eyewear equipment for the user based on the generated user eyewear equipment profile.