Abstract: A method, apparatus and computer program product are provided for learning to generate maps from raw geospatial observations from sensors traveling within an environment.

Abstract: A system receives information associated with an interaction with an individual in a context. Then, the system analyzes the information to extract features associated with one or more attributes of the individual. Moreover, the system generates, based at least in part on the extracted features, a group of behavioral agents in a multi-layer hierarchy that automatically mimics the one or more attributes. Next, the system calculates one or more performance metrics associated with the group of behavioral agents and the one or more attributes. Furthermore, the system determines, based at least in part on the one or more performance metrics, one or more deficiencies in the extracted features. Additionally, the system selectively acquires second information associated with additional interaction with the individual in the context based at least in part on the one or more deficiencies to at least in part correct for the one or more deficiencies.

Abstract: In accordance with one embodiment of the present disclosure, a method includes receiving a set of images, each image depicting a view of a scene, generating sparse depth data from each image of the set of images, training a monocular depth estimation model with the sparse depth data, generating, with the trained monocular depth estimation model, depth data and uncertainty data for each image, training a NeRF model with the set of images, wherein the training is constrained by the depth data and uncertainty data, and rendering, with the trained NeRF model, a new image having a new view of the scene.

Abstract: Techniques are described for bone graft selection in orthopedic surgery. Processing circuitry may determine a bone graft template for a bone graft to be connected to a first anatomical object, determine information indicative of placement of the bone graft template within a representation of a second anatomical object based on image data for one or more images of anatomical objects, and output the information indicative of the placement of the bone graft template within the representation of the second anatomical object.

Abstract: Systems and techniques are provided for enhancing operations of a ray tracing processor. For instance, a process can include obtaining one or more nodes of an acceleration data structure. Each node of the one or more nodes includes the same number of bytes. The node(s) can be stored in a cache associated with a ray tracing processor. Each of the stored node(s) are cache line-aligned with the cache associated with the ray tracing processor. A first stored node of the stored node(s) can be provided to the ray tracing processor and processed by the ray tracing processor during a first clock cycle of the ray tracing processor. A second stored node of the stored node(s) can be provided to the ray tracing processor and processed by the ray tracing processor during a second clock cycle of the ray tracing processor.

Abstract: A keypoint detection system includes: a camera system including at least one camera; and a processor and memory, the processor and memory being configured to: receive an image captured by the camera system; compute a plurality of keypoints in the image using a convolutional neural network including: a first layer implementing a first convolutional kernel; a second layer implementing a second convolutional kernel; an output layer; and a plurality of connections between the first layer and the second layer and between the second layer and the output layer, each of the connections having a corresponding weight stored in the memory; and output the plurality of keypoints of the image computed by the convolutional neural network.

Abstract: Various implementations disclosed herein include devices, systems, and methods that render perspective correct 2D graphics. In some implementations, a vector graphic including portions defined on a 2D canvas is obtained, and tiles of the 2D canvas is divided into tiles. In some implementations, a subset of the portions outside of a first tile of the tiles on the 2D canvas is identified. Then, the portions of the subset are projected onto an edge of the first tile. In some implementations, data for rendering the vector graphic is provided, the data comprising the projected portions of the subset of the portions outside the first tile.

Abstract: Mechanisms for image processing are provided. A computing device comprises a memory, an input device, a display, and a processor. The processor is configured to: acquire a three-dimensional image of an anatomical structure and store it in the memory. The processor renders on the display (i) an initial volume of the three-dimensional image corresponding to an initial portion of the anatomical structure, and (ii) a moveable control element. The initial volume has an outer surface defined by a position of the control element. The processor receives input data updating the position of the control element relative to the initial volume; and renders on the display, in place of the initial volume, a further volume of the three-dimensional image, corresponding to a further portion of the anatomical structure and having a further outer surface defined by the updated position of the control element.

Abstract: A method for facilitating wave-based inversion using Time-Domain Sifrian Inversion (TDSI). The method includes receiving, using a communication device, observed data from a device. The method includes obtaining, using a processing device, an initial model. The method includes obtaining, using the processing device, synthetic data from the initial model. The method includes determining, using the processing device, a matching of the synthetic data with the observed data is not within a matching threshold. The method includes updating, using the processing device, the initial model using the TDSI iteratively until the matching threshold for the matching is achieved. The TDSI provides an update for the updating of the initial model iteratively using a time-domain cost function which is robust to a cycle skipping. The method includes generating, using the processing device, a final model based on the updating. The method includes storing, using a storage device, the final model.

Abstract: Novel viewpoints may be rendered using arrays of camera images with synthesized viewpoints. Different viewpoints of a scene may be captured via image sensors. Depth information may be determined for the captured viewpoints. An array of images may be generated that includes synthetic viewpoints generated from the captured viewpoints and depth information. A request to render a novel viewpoint may be received. Then novel viewpoint may be rendered using the array of images with synthetic viewpoints.

Abstract: A method for generating a texture map used during a video conference, the method may include obtaining multiple texture maps of multiple areas of at least a portion of a three-dimensional (3D) object in which the multiple texture maps comprise a first texture map of a first area and of a first resolution, and a second texture map of a second area and of a second resolution, in which the first area differs from the second area and the first resolution differs from the second resolution; generating a texture map of the at least portion of the 3D object. The generating is based on the multiple texture maps; and utilizing the visual representation of the at least portion of the 3D object based on the texture map of the at least portion of the 3D object during the video conference.

Abstract: The present disclosure describes approaches to camera re-localization that improve the accuracy of re-localization determinations by performing simulated consistency checks for three-dimensional maps. Client devices associated with users of a location-based application transmit image scans to a game server, which divides the received scan data into mapping sets used to generate 3D maps of environments and validation sets used to test the accuracy of the maps. To perform the testing, the game server identifies query scans in the validation set having GPS coordinates within a threshold distance of the mapped location and uses the 3D map of the environment to generate a pose estimate for each frame. The results of the localization queries are analyzed by comparing differences between the localization pose estimates and differences between the poses of independent pairs of frames in the query scan to evaluate the accuracy of the 3D map.

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 cricket sensor may include one or more first image-capturing sensors configured to capture image data of a pitching motion of a bowler and image data of an initial motion of a cricket ball at a bowling end of a cricket field. The cricket sensor may include one or more second image-capturing sensors configured to capture image data of a trajectory and a flight path of the cricket ball towards a batting end of the cricket field. The cricket sensor may also include one or more first radar sensors configured to capture radar data describing one or more initial launch parameters of the cricket ball related to the trajectory and the flight path of the cricket ball towards the batting end of the cricket field.

Abstract: The present disclosure relates generally to a system and method for a user to control a virtual representation of themselves within a three-dimensional virtual world. The system and method enable utilizing a two-dimensional image or video data of user with extracted depth information to position themselves in a three-dimensional scene. It also provides a control system and method for a user to control the virtual representation of themselves using the output video as a visual feedback mechanism in a three-dimensional space including the virtual representation of themselves. A user interacts with other virtual objects or items in a scene or even with other users visualized in the scene.

Abstract: The present disclosure provides new and innovative systems and methods for generating eye models with realistic color. In an example, a computer-implemented method includes obtaining refraction data, obtaining mesh data, generating aligned model data by aligning the refraction data and the mesh data, calculating refraction points in the aligned model data, and calculating an approximated iris color based on the refraction points and the aligned model data by calculating melanin information for the aligned model data based on the refraction points for iris pixels in the aligned model data.

Abstract: A method for optimizing a photographing pose of a user, where the method is applied to an electronic device, and the method includes: displaying a photographing interface of a camera of the electronic device; obtaining a to-be-taken image in the photographing interface; determining, based on the to-be-taken image, that the photographing interface includes a portrait; entering a pose recommendation mode; and presenting a recommended human pose picture to a user in a predetermined preview manner, where the human pose picture is at least one picture that is selected from a picture library through metric learning and that has a top-ranked similarity to the to-be-taken image, and where the similarity is an overall similarity obtained by fusing a background similarity and a foreground similarity.

Abstract: An apparatus, method and computer program are disclosed. The apparatus may include circuitry configured for determining, relative to one or more real audio sources in an environment, a relative position of a virtual object in the environment. The circuitry may be configured for determining, based on the relative position, an occluded region of the environment in which audio signals from the one or more real audio sources would be occluded by the virtual object. The circuitry may be configured for outputting, to a user in the environment, audio data, the outputting circuitry configured to operate in a first mode and in a second mode. The circuitry may be configured for switching the outputting means circuitry between the first mode and the second mode based on a location of the user relative to the occluded region.

Abstract: A method and system for hybrid streaming of media content such as video for a computer-based game. In some implementations, a background media content stream includes a background video rendered on a server. The background media content stream is received at a client device. The client device renders one or more three-dimensional objects and forms a composite media stream by combining the background media content stream and the one or more three-dimensional objects. The composite media stream can be output to a display or other external system.

Abstract: A method, apparatus and system for portfolio performance prediction of a portfolio of projects include receiving information regarding at least start date delays, duration increases and cost overruns for at least one project of the portfolio of projects, determining at least one cluster for the at least one project from the received information regarding at least the start date delays, duration increases and cost overruns for the at least one project of the portfolio of projects, creating a statistical representation for each of the clusters of the at least one project of the portfolio, and predicting a performance of the portfolio of projects using information regarding the statistical representation of the clusters of the at least one project of the portfolio of projects.

Abstract: To easily and quickly detect a predetermined shape on an object. A shape detection method for detecting the shape of an object, the method comprising: a step of acquiring data indicating three-dimensional point group data from the object by a sensor, a step of displaying the three-dimensional point group data in a display area of a terminal, a step of acquiring the display area in which the three-dimensional point group data is displayed, as two-dimensional image data, and a step of performing image analysis on the two-dimensional image data, and detecting a predetermined shape of the object.

Abstract: There is a region of interest of a synthetic image depicting an object from a class of objects. A trained neural image generator, having been trained to map embeddings from a latent space to photorealistic images of objects in the class, is accessed. A first embedding is computed from the latent space, the first embedding corresponding to an image which is similar to the region of interest while maintaining photorealistic appearance. A second embedding is computed from the latent space, the second embedding corresponding to an image which matches the synthetic image. Blending of the first embedding and the second embedding is done to form a blended embedding. At least one output image is generated from the blended embedding, the output image being more photorealistic than the synthetic image.

Abstract: An imaging system including a first camera and a second camera corresponding to a first eye and a second eye of a user, respectively; and at least one processor. The at least one processor is configured to control the first camera and the second camera to capture a sequence of first images and a sequence of second images of a real-world environment, respectively; and apply a first extended depth-of-field correction to one of a given first image and a given second image, whilst applying at least one of: defocus blur correction, image sharpening, contrast enhancement, edge enhancement to another of the given first image and the given second image.

Abstract: Method of generating modified media content item based on user's setting selection starts with a processor causing a camera personalization interface to be displayed by a display of a client device. The processor receives a setting selection from the client device via the camera personalization interface. The processor determines a pre-capture setting and a post-processing setting based on the setting selection. The processor calibrates a camera of the client device using the pre-capture setting. The processor receives a media content item including an image captured using the camera and generates a modified media content item by modifying the media content item using the post-processing setting. Other embodiments are described herein.

Abstract: Methods and systems for spectating a live video game are presented. In one method embodiment, an operation includes providing an interface for presenting an overhead view of an interactive environment of a video game on a head mounted display (HMD). The method also provides an operation for providing a magnifying window for displaying a magnified view of the interactive environment, where the magnified view is associated with a second vantage point that is closer to the interactive environment than is a first vantage point of the overhead view. The method further provides an operation for tracking a real-world position of a controller of the spectator and moving the location of the magnifying window to correspond to the real-world position of the controller.

Abstract: Disclosed is a graphics system and associated methods for preserving or improving image quality when rendering a decimated image or three-dimensional (“3D”) that has been decimated to remove some primitives from the original representation of the undecimated image or 3D model. The graphics system receives decimated primitives that are defined with a position, visual characteristics, and at least first and second surface normals. The graphics system defines a light source for illuminating the decimated primitives, determines that the second surface normal of a particular decimated primitive receives more light from the light source than the first surface normal of the particular decimated primitive, and generates a visualization for the particular primitive at the position of the particular primitive with the visual characteristics of the particular primitive adjusted according to an amount of light from the light source reaching the particular primitive via the second surface normal.

Abstract: A location-based metaverse social media system provided for users to actually move and carry out various social events of online/offline interactions after positioning through a mobile device includes a virtual world management processor and an user processor. The virtual world management processor is provided for augmenting a plurality of parallel virtual worlds by basic map information of the real world and the user processor is provided for the users to create personal accounts and a virtual character tied with them and displayed on the mobile device. In operations, the users can switch to any one of the parallel virtual worlds in a cross-world manner and perform a location-based social event while the users are moving.

Abstract: Systems and methods for three-dimensional object category modeling can utilize figure-ground neural radiance fields for unsupervised training and inference. For example, the systems and methods can include a foreground model and a background model that can generate an object output based at least in part on one or more learned embeddings. The foreground model and background model may process position data and view direction data in order to output color data and volume density data for a respective position and view direction. Moreover, the object category model may be trained to generate an object output, which may include an instance interpolation, a view synthesis, or a segmentation.

Abstract: A computer implemented method of navigating a digital representation of a spatial environment by a user in which a portion of the digital representation is rendered by a rendering computer system for visualization by the user. The method including accessing a digital model of the spatial environment including a specification of each of a plurality of virtual resources in the spatial environment, each specification identifying at least a location of a resource in the spatial environment and at least one trigger criterion for determining when a resource is accessed by the user; monitoring the satisfaction of trigger criteria for the plurality of virtual resources; and responsive to the monitoring, determining the portion of the digital representation for rendering by the rendering computer system for visualization by the user.

Abstract: A system for building information modeling comprising a plurality of 2D documentation sets associated with a building. A 3D model associated with the building, the 3D model including a plurality of user-selectable controls, wherein each user-selectable control comprises an icon and an associated balloon that is generated when the icon is selected. The balloon for each user-selectable control comprising a first selection control that causes one of the plurality of 2D documentation sets to be shown after an animation sequence. The balloon for each user-selectable control comprising a second selection control that causes a transition to a 3D model having an overlay of the 2D documentation sets.

Abstract: Methods, devices, and systems related to a computing device for capturing and displaying an augmented reality (AR) image are described. An example method can include capturing a first three-dimensional (3D) figure of a user at a mobile device. The method further includes accessing the first 3D figure of a user at the mobile device. The method further includes receiving, at the mobile device, a selection of a product to display on the 3D figure. The method further includes displaying, on the mobile device, the product to display on the first 3D figure as an augmented reality (AR) image.

Abstract: A method for determining a collision range includes: determining a grid plane corresponding to a model, the grid plane including a plurality of grid regions; projecting the model to the grid plane and obtaining height information of the model with respect to the grid plane, the height information of the model including grid height information corresponding to a grid region that identifies a height of a sub-model in the model with respect to the grid plane, and the sub-model being a portion of the model in a projection space of the grid region; and obtaining a collision range of the model according to the grid plane and the height information.

Abstract: A system including a processor configured together and analyzing network activity data associated with a plurality of network service providers, gather geographic reference data that includes information about locale(s) of interest, create bins and use spatial matching to associate network performance parameters with hex bins associated with the locale(s) of interest. Generate, based at least in part on the geographic network data, a geospatial map overlay for incorporation in an interactive graphical user interface (GUI), the geospatial map overlay comprising a visual indication of one or more areas of interest and associated aggregated network performance data, and identify and deploy one or more measures to improve network performance and thus customer experiences for at least one or more of the identified locale associated with a bin.

Abstract: In method for replaying a scanning process according to the present invention, position information and rotation information of a scanner are acquired while scanning a scan target is a scanning step, and the distance relationship and angle relationship with the scan target can be calculated according to the position information and rotation information of the scanner. The acquired information and relationship values can be linked with a scan time and stored, and scanning data, along with the acquired information, can be sequentially displayed over time in a display step. According to the method for replaying a scanning process, the process of scanning performed by a user can be examined in chronological order, and the examination process can be quickly performed by controlling replay speed and selecting scan time.

Abstract: Methods and control stream generators for generating a control stream for a tile group comprising at least two tiles, the control stream identifying primitive blocks that are relevant to rendering at least one tile in the tile group.

Abstract: A system for updating continuous image alignment of separate cameras identifies a previous alignment matrix associated with a previous frame pair captured at one or more previous timepoints by a reference camera and a match camera. The previous alignment matrix is based on visual correspondences in the previous frame pair. The system also identifies a current matrix associated with a current frame pair captured at one or more current timepoints by the reference camera and the match camera. The current matrix is based on visual correspondences in the current frame pair. The system also identifies a difference value associated with the reference camera or the match camera relative to the one or more previous timepoints and the one or more current timepoints. The system also generates an updated alignment matrix by using the previous alignment matrix, the current matrix, and the difference value as inputs.

Abstract: Systems, apparatus, and methods for rendering content based on display assembly pose are disclosed. In an example, motion capture data of a display of a plurality of displays is received, the display moving from a first physical pose to a second physical pose. The motion capture data is processed to determine the coordinates of the second physical pose. A transformation of the second physical pose of the display to a virtual pose of the display is generated. A virtual model of the plurality of displays is updated, the virtual model comprising the virtual pose of the display. The content is rendered on the updated virtual model.

Abstract: A method and process for providing three dimensional (3-D) virtual image of a patient in metaverse for surgical or other procedures, wherein the avatar of a medical practitioner, imitating the actions of the practitioner performing the procedure, can visually identify the organs and the location of the instruments in real-time inside patient. Such an image reconstruction with spatial coordination provides a usable metaverse implementation with the medical professional's persona as avatar, usable as a training and supportive tool in medical applications. It is a very valuable, especially for the medical and surgical community. The implemented Metaverse provides all details collected and combined from the multiple imaging systems. It is usable as a diagnostic tool, a practice tool a teaching tool and real time direction and feedback tool by the medical community for procedures. The real-time image implemented as Metaverse provides critical visual capabilities during procedures.

Abstract: The system which dynamically determines crawl frequency. Crawl frequency is computed by determining the proportional integral derivative (PID) contribution of certain data to the system. Alternatively, the system determines the next time to crawl based on both the rate of change and displacement of data. Maximum and minimum wait times between crawls are taken into account for the computation. In another embodiment, crawl frequency is calculated based on the rate of change. The system determines the next time to crawl based on the rate of change of data within set parameters, such as maximum and minimum wait times between crawls. Alternatively, the system implements a recurrent neural network (RNN) using long short-term memory (LSTM) units for a future data prediction. With this information, the system determines the next time to crawl.

Abstract: The present invention provides a system for promoting health equity in diverse and underserved groups. It targets equity-affected diseases where disadvantaged groups face disproportionate clinical outcomes, such as, for example, peanut and tree nut food allergies, hypertension, diabetes, maternal mortality and sepsis, and adjusts for biases and inequity disease-causing factors using artificial intelligence systems. Multifaceted financial systems, investor segmentation, and health data profiling facilitate strategic investments. The system of this invention promotes ongoing monitoring, evaluation, and subsequent investment outreach to a health equity funding pool, fostering comprehensive health equity initiatives for diverse and underserved groups while prioritizing data security.

Abstract: A tracking system and method are provided. At least one tracking device receives a first role configuration from a head-mounted display. The at least one tracking device determines at least one first spatial area from a spatial map information stored in a storage of the at least one tracking device based on the first role configuration. The at least one tracking device loads the at least one first spatial area into a memory of the at least one tracking device.

Abstract: A vehicle interior component including a patterned padding surface positioned proximate to an occupant of the vehicle. The component includes a substrate layer with a plurality of padding projections and an exterior layer connected to the substrate layer and including a plurality of gaps matching the location of the substrate layer's plurality of padding projections. Each of the plurality of padding projections aligns with one of the plurality of gaps and is exposed above the exterior layer to thereby form a pattern of padding on the surface of the vehicle interior component.

Abstract: A method includes a computing system associated with a first artificial-reality device accessing sensor data of a real environment, generating a first plurality of feature descriptors for the real environment, and generating a second plurality of feature descriptors for the real environment based on the first plurality of feature descriptors, wherein the second plurality of feature descriptors have lower detail than the first plurality of feature descriptors. The computing system may further cause the first plurality of feature descriptors to be transmitted using a wireless connection, wherein the transmitted first plurality of feature descriptors are configured to be converted into a third plurality of feature descriptors different from the second plurality of feature descriptors. The third plurality of feature descriptors are configured to be used by a second artificial-reality device for tracking the real environment.

Abstract: A method for processing point cloud data according to embodiments may encode and transmit point cloud data. The method for processing point cloud data according to embodiments may receive and decode point cloud data.

Abstract: An electronic device for discharging a pulse of current includes an energy storage medium for receiving and storing electric energy supplied from a power source, a solenoid and a control switch connectable with the energy storage medium and the power source and configured to discharge a pulse of current for energising the solenoid when the control switch is activated, whereby the solenoid moves from a first position when the solenoid is not energised, into a second position when the solenoid is energised, and the energised solenoid generates and supplies a current back to the energy storage medium, a processor configured to process the time taken to energise the solenoid, monitor the current drawn by the solenoid from the energy storage medium, and control at least one parameter of a discharged pulse; and an output module to control the discharge of the current from the energy storage medium to the solenoid.

Abstract: A computing system is provided comprising a processor is configured to execute an augmented reality avatar program to receive a second-user avatar-data object and second-user pose information, via a computer network, receive an image from a live image feed captured by a camera of the computing device, identify a virtual surface corresponding to a physical surface that appears in the image; calculate a position and a pose of a second-user avatar relative to the virtual surface based on the second-user pose information, render the second-user avatar-data object to thereby display the second-user avatar of the second user superimposed upon the image. The second-user avatar is anchored to the virtual surface according to the calculated position and pose of the second-user avatar. The virtual surface is world-locked to the physical surface in the image in a real world 3-D environment.

Abstract: This disclosure relates to methods, non-transitory computer readable media, and systems that can render a virtual object in a digital image by using a source-specific-lighting-estimation-neural network to generate three-dimensional (“3D”) lighting parameters specific to a light source illuminating the digital image. To generate such source-specific-lighting parameters, for instance, the disclosed systems utilize a compact source-specific-lighting-estimation-neural network comprising both common network layers and network layers specific to different lighting parameters. In some embodiments, the disclosed systems further train such a source-specific-lighting-estimation-neural network to accurately estimate spatially varying lighting in a digital image based on comparisons of predicted environment maps from a differentiable-projection layer with ground-truth-environment maps.

Abstract: A point cloud data transmission method according to embodiments comprises the steps of: acquiring point cloud data; encoding geometry information included in the point cloud data; encoding attribute information included in the point cloud data; and transmitting the encoded geometry information and the encoded attribute information.

Abstract: Systems and methods for processing geographical zones are provided. The method involves operating at least one processor to: define a bounding box surrounding a geographical zone; iteratively partition the bounding box into a plurality of bounding boxes by, starting with the bounding box in the first iteration: determine whether the bounding box contains more than a predetermined maximum number of vertices; divide the bounding box into two additional bounding boxes if the bounding box contains more than the predetermined maximum number of vertices, repeat the steps of determining and dividing the bounding box for the additional bounding boxes in the next iteration until each additional bounding box does not contain more than the predetermined maximum number of vertices; generate a binary tree data structure representing the geographical zone; search the binary tree data structure to determine whether a vehicle is located within one of the bounding boxes.

Abstract: A method includes obtaining an image of a damaged vehicle and a three-dimensional (3D) vehicle model associated with the damaged vehicle. The method also includes identifying one or more damage markers that indicate damage to the damaged vehicle based on the image. The method further includes identifying a pose of the damaged vehicle within the image. In addition, the method includes mapping the one or more damage markers onto the 3D vehicle model based on the identified pose, where the one or more mapped damage markers identify one or more locations and dimensions of the damage to the damaged vehicle on the 3D vehicle model. In some cases, one or more trained machine learning models may be used to identify the one or more damage markers, identify the pose of the damaged vehicle within the image, and/or map the one or more damage markers onto the 3D vehicle model.