Abstract: In implementations of systems for generating candidate mirror snap points using determined axes of symmetry, a computing device implements a symmetry system to receive vector object data describing a set of points of a vector object. The symmetry system generates convex polygons that enclose the set of points and identifies a particular convex polygon that has a smallest area. A side of the particular convex polygon is determined as an axis of symmetry for the vector object. The symmetry system generates an indication for display in a user interface of a candidate snap point based on the axis of symmetry and a point of the set of points of the vector object.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer-readable media that generate interactive visual shape representation of digital datasets. For example, the disclosed systems can generate an augmented nearest neighbor network graph from a sampled subset of digital data points using a nearest neighbor model and witness complex model. The disclosed system can further generate a landmark network graph based on the augmented nearest neighbor network graph utilizing a plurality of random walks. The disclosed systems can also generate a loop-augmented spanning network graph based on a partition of the landmark network graph by adding community edges between communities of landmark groups based on modularity and to complete community loops. Based on the loop-augmented spanning network graph, the disclosed systems can generate an interactive visual shape representation for display on a client device.

Abstract: Glyph selection techniques are described for digital text content that enable selection of non-contiguous glyphs via input describing at least one path of any shape or form relative to the digital text content. A text selection system receives freeform drawing input defining a path and ascertains a selection scope for outputting a text selection relative to the path. The selection scope indicates whether the text selection is to include glyphs that intersect the path, glyphs displayed within an area enclosed by the path, glyphs displayed outside an area enclosed by the path, or combinations thereof. The text selection system is configured to identify a subset of glyphs for inclusion in the text selection, without constraint as to whether the subset of glyphs are contiguous in the digital text content, and output the subset of glyphs as a single selection of the digital text content.

Abstract: Vector object stylization techniques from raster objects are described that support editing of vector objects in a manner that maintains an underlying mathematical representation of object. A raster object, for instance, is generated from an edited version of an output of a vector object. This raster object, along with the vector object are received as inputs by a vector conversion system. These inputs are utilized by the vector conversion system to generate a stylized vector object having a visual appearance that mimics and simulates a visual appearance of the raster object. As a result, the stylized vector object provides a mathematical representation of the raster object.

Abstract: A computer implemented method of generating at least one anonymous image, comprises: extracting and preserving at least one real facial region from at least one real image of a real human face, and generating at least one anonymous image comprising a synthetic human face and the preserved at least one real facial region, wherein an identity of the real human face is non-determinable from the at least one anonymous image.

Abstract: A method for computer animation includes receiving, by a processor executing a multi-platform deformation engine system in association with a third-party game engine system, an input file that includes an asset geometry, where the asset geometry defines an asset mesh structure, where the asset geometry may exclude an internal support frame, and where logic for custom deformation steps may be included, altogether in a fashion portable and made to produce consistent results across multiple different software and/or hardware platform environments and/or across real-time and/or offline scenarios. The method also includes applying at least one deformer to the asset mesh structure, where the at least one deformer includes a plurality of user-selectable deformer channels, and where each deformer channel is associated with at least a portion of the asset mesh structure and configured to adjust a visual appearance of the associated portion of the asset mesh structure.

Abstract: A fully-connected neural network may be configured for execution by a processor as a fully-fused neural network by limiting slow global memory accesses to reading and writing inputs to and outputs from the fully-connected neural network. The computational cost of fully-connected neural networks scale quadratically with its width, whereas its memory traffic scales linearly. Modern graphics processing units typically have much greater computational throughput compared with memory bandwidth, so that for narrow, fully-connected neural networks, the linear memory traffic is the bottleneck. The key to improving performance of the fully-connected neural network is to minimize traffic to slow “global” memory (off-chip memory and high-level caches) and to fully utilize fast on-chip memory (low-level caches, “shared” memory, and registers), which is achieved by the fully-fused approach.

Abstract: Disclosed is surface guided cropping in volume rendering of 3D volumetric data from intervening anatomical structures in the patient's body. A digital 3D representation expressing the topography of a first anatomical structure is used to define a clipping surface or a bounding volume which then is used in the volume rendering to exclude data from an intervening structure when generating a 2D projection of the first anatomical structure.

Abstract: The present disclosure relates to methods and apparatus for graphics processing. Aspects of the present disclosure can determine at least one scene including one or more viewpoints. Also, aspects of the present disclosure can divide the at least one scene into a plurality of zones based on each of the one or more viewpoints. Further, aspects of the present disclosure can determine whether a zone based on one viewpoint of the one or more viewpoints is substantially similar to a zone based on another viewpoint of the one or more viewpoints. Aspects of the present disclosure can also generate a geometry buffer for each of the plurality of zones based on the one or more viewpoints. Moreover, aspects of the present disclosure can combine the geometry buffers for each of the plurality of zones based on the one or more viewpoints.

Abstract: Forming a 3D representation of an object using a portable electronic device with a camera. A sequence of image frames, captured with the camera may be processed to generate 3D information about the object. This 3D information may be used to present a visual representation of the object as real-time feedback to a user, indicating confidence of 3D information for regions of the object. Feedback may be a composite image based on surfaces in a 3D model derived from the 3D information given visual characteristics derived from the image frames. The 3D information may be used in a number of ways, including as an input to a 3D printer or an input, representing, in whole or in part, an avatar for a game or other purposes. To enable processing to be performed on a portable device, a portion of the image frames may be selected for processing with higher resolution.

Abstract: A computer simulation object such as a chair is described by voice or photo input to render a 2D image. Machine learning may be used to convert voice input to the 2D image. The 2D image is converted to a 3D asset and the 3D asset or portions thereof are used in the computer simulation, such as a computer game, as the object such as a chair.

Abstract: The present disclosure relates to methods and apparatus for graphics processing. The present disclosure can calculate a center-edge distance of a first pixel, the center-edge distance of the first pixel equal to a distance from a first pixel center to one or more edges of a first primitive. Additionally, the present disclosure can store the center-edge distance of the first pixel when the first primitive is visible in a scene. The present disclosure can also determine an amount of overlap between the first pixel and the first primitive. Further, the present disclosure can blend a color of the first pixel with a color of a second pixel based on at least one of the center-edge distance of the first pixel or the amount of overlap between the first pixel and the first primitive.

Abstract: A method of rendering an image of an environment is disclosed. Environment data for the environment is accessed. The environment data corresponds to a frame of a video. A plurality of subframes associated with the frame is determined. An angle for each of the plurality of subframes is determined. One or more lights corresponding to the environment are selected. For each light of the one or more lights, a shadow map is generated. The shadow map corresponds to a subframe of the plurality of subframes based on a frustum view oriented at the angle determined for the subframe. The image of the environment is rendered. The rendering includes using the generated shadow map for each light of the one or more lights.

Abstract: Embodiments of the present application provide a data processing method and an electronic device. The data processing method includes: determining whether a current collection scene satisfies a condition for enabling a high-dynamic range (HDR) collection function; automatically enabling the HDR collection function in response to the current collection scene satisfying the condition for enabling the HDR collection function; and collecting at least two two-dimensional images with different exposures within a collection time of one frame of three-dimensional video data based on the HDR collection function; wherein the at least two two-dimensional images are configured to enable a mobile edge computing (MEC) server to build a three-dimensional video.

Abstract: Techniques of compressing triangular mesh data involve generating a neighborhood table (i.e., a table) of fixed size that represents a neighborhood of a predicted vertex of a triangle within a triangular mesh for input into a machine-learning (ML) engine. For example, such a neighborhood table as input into a ML engine can output a prediction for a value (e.g., a position) of a vertex. The residual between the prediction and the actual value of the vertex is stored in an array. The data in the array representing the residuals may be compressed and transmitted over a network. Upon receipt by a computer, the array may be decompressed by the computer. Obtaining the actual value involves the receiving computer generating the same neighborhood table, inputting that neighborhood table into the same ML engine to produce the predicted value, and adding the predicted value to the residual from the decompressed file.

Abstract: A method for generating a three-dimensional working surface of a human body includes receiving input data corresponding to images; generating a first point cloud from the input data, each point being associated with a three-dimensional spatial coordinate; determining a plurality of attributes at each point of the first point cloud, calculating a set of geometric parameters from a regression carried out from a series of matrix operations performed according to different layers of a neural network trained from a plurality of partial views of parametric models parameterised with different parameterisation configurations; determining a parameterised model to generate a body model of a body including a first meshing.

Abstract: Digital object surface inflation techniques are described as implemented by an image processing system to generate an inflated digital object, automatically and without user intervention, from a two-dimensional digital object in a digital image. In one example, the geometry inflation system generates the inflated digital object based on an outer boundary and in this way overcomes challenges of conventional skeleton based techniques. In another example, rules are configured to guide a geometry inflation system to generate the inflated digital object.

Abstract: A computer-implemented method of augmented reality includes capturing the video flux with a video camera, extracting, from the video flux, one or more 2D images each representing the real object, and obtaining a 3D model representing the real object. The method also includes determining a pose of the 3D model relative to the video flux, among candidate poses. The determining rewards a mutual information, for at least one 2D image and for each given candidate pose, which represents a mutual dependence between a virtual 2D rendering and the at least one 2D image. The method also includes augmenting the video flux based on the pose. This forms an improved solution of augmented reality for augmenting a video flux of a real scene including a real object.

Abstract: A context based augmented reality system can be used to display augmented reality elements over a live video feed on a client device. The augmented reality elements can be selected based on a number of context inputs generated by the client device. The context inputs can include location data of the client device and location data of nearby physical places that have preconfigured augmented elements. The preconfigured augmented elements can be preconfigured to exhibit a design scheme of the corresponding physical place.

Abstract: Systems and methods for conveying virtual content in an augmented reality environment comprising images of virtual content superimposed over physical objects and/or physical surroundings visible within a field of view of a user as if the images of the virtual content were present in the real world. Exemplary implementations may: obtain user information for a user associated with a presentation device physically present at a location of the system; compare the user information with the accessibility criteria for the virtual content to determine whether any portions of the virtual content are to be presented to the user based on the accessibility criteria and the user information for the user; and facilitate presentation of the virtual content to the user via presentation device of user based on the virtual content information, the field of view, and the correlations between the multiple linkage points and the reference frame of the virtual content.

Abstract: A method, according to one example, includes receiving, by a system including a processor, a three-dimensional (3D) data array defined in a 3D rectangular coordinate space. The method includes transforming, by the system, the 3D data array into a radial array defined in a spherical coordinate space. The method includes generating, by the system, a 3D immersive visualization of the radial array in which a user appears to be immersed within the radial array. The method includes continually modifying the 3D immersive visualization based on movement of the user.

Abstract: Graphical style modification may be implemented using machine learning. A color accommodation module receives an image frame from a host system and generates a color-adapted version of the image frame. A Graphical Style Modification module generates a style adapted video stream by applying a style adapted from a target image frame to each image frame in a buffered video stream.

Abstract: Disclosed herein is a method of graphically presenting an indicating marker over a 3-D model of a tissue surface during a catheterization procedure, comprising determining a region over the 3-D model, deforming the indicating marker to congruently match a shape defined by the 3-D model across the region at a plurality of positions; and rendering the 3-D model into an image including the deformed indicating marker by generating an image of the 3-D model covered by said deformed indicating marker.

Abstract: Methods and devices for manipulating an image are described. The method comprises receiving image data, the image data including a first image obtained from a first camera and a second image obtained from a second camera, the first camera and the second camera being oriented in a common direction; identifying one or more boundaries of an object in the image data by analyzing the first image and the second image; and displaying a manipulated image based on the image data, wherein the manipulated image includes manipulation of at least a portion of the first image based on boundaries of the object.

Abstract: An information board is displayed and used in a three-dimensional virtual reality environment wherein the information board is extended by adding discrete board segments through control inputs received from a user in the virtual reality environment.

Abstract: The present disclosure illustrates systems and methods for automatically adjusting a following 3D asset based on a deformation of a related base 3D asset. The systems and methods may use geomaps to index the relationship between the following 3D asset and base 3D asset. By automatically adjusting a following 3D asset based on the base 3D asset, the following 3D asset may retain full functionality.

Abstract: The present disclosure relates to a method, device, system and computer-program product for setting a lighting condition when an object is checked and a storage medium. The method includes that: the object is lighted by light sources capable of changing lighting parameters, and the object is captured by an image sensor in such lighting parameters to obtain captured images, wherein the object has known label data; and a part of or all of the captured images and the corresponding label data of the object are applied to learning of a machine learning model, and the lighting condition and the check algorithm parameters of the machine learning model is set simultaneously by optimizing both the lighting parameters and the check algorithm parameters, on the basis of a comparison result between an estimation result of the machine learning model and the label data. Therefore, operations are simplified.

Abstract: Image capture devices for monitoring parking at various geographic locations and systems for monitoring parking are described herein.

Abstract: A chip recognition system in which a chip is configured to at least partially have a specific color indicative of a value of the chip includes: a recording device that uses a camera and records a state of the chip as an image; an image analysis device that subjects the image so recorded to image analysis and recognizes at least the specific color and a reference color that is present in the image and differs from the specific color; and a recognition device at least including an artificial intelligence device that uses a result of the image analysis by the image analysis device and specifies the specific color of the chip, wherein the artificial intelligence device of the recognition device has been subjected to teaching using, as training data, a plurality of images of the chip and the reference color irradiated with different illumination intensities.

Abstract: Variation in received documents types and templates used for each document type poses challenge in developing a generic background noise removal approach for automatic text information extraction technique. Embodiments herein provide a method and a system for document classification and text information extraction. Time efficient and accurate text detection engine-based Region of Interest (ROI) technique is provided to accurately identify text region followed by a multi-layered neural network based architecture for enhanced classification accuracy to identify the type of document. A multistage image pre-processing approach is provided for efficient, effective, and accurate background noise removal from the classified document, which includes unsupervised clustering, identification, segmentation, masking, contour approximation, selective subtraction, and dynamic thresholding.

Abstract: The present disclosure relates to an object selection system that accurately detects and optionally automatically selects user-requested objects (e.g., query objects) in digital images. For example, the object selection system builds and utilizes an object selection pipeline to determine which object detection neural network to utilize to detect a query object based on analyzing the object class of a query object. In particular, the object selection system can identify both known object classes as well as objects corresponding to unknown object classes.

Abstract: In variants, the method for occlusion correction can include: determining a measurement depicting an occluded object of interest (OOI), optionally infilling the occluded portion of the object of interest within the measurement, and determining an attribute of the object of interest based on the infilled measurement.

Abstract: An apparatus for contextual execution comprises a processor, and a memory containing instructions, which when executed by the processor, cause the apparatus to receive, from a user terminal, a control input associated with an intent, obtain location data associated with a location of the user terminal, and determine a scored set of execution options associated with the control input. Further, the instructions, when executed by the processor cause the apparatus to obtain a contextual label associated with the location data, the label determined based on the application of one or more adapted pretrained deep learning models to the location data.

Abstract: Disclosed is an infrared thermal image classification and hot spot positioning method of a photovoltaic panel, comprising following steps: constructing a photovoltaic panel infrared thermal image data set, preprocessing the image data set, and dividing into a training set and a testing set according to a preset proportion; training an auxiliary generating countermeasure network based on the training set to obtain a trained generator and a trained discriminator, training an encoder by combining the image data in the training set with the generator, and fixing the parameters of a trained encoder; inputting the image data in the test set into a trained discriminator to obtain a photovoltaic image classification result; inputting images in the test set into the trained encoder, and then inputting the images into the generator for a reconstruction; comparing pixels of the input images with the reconstructed infrared thermal images, and positioning hot spots.

Abstract: Disclosed is a method for recognizing distribution network equipment based on Raspberry Pi multi-scale feature fusion. The method includes obtaining an initial sample data set; constructing an object detection network composed of EfficientNe-B0 backbone network, multi-scale feature fusion module and a regression classification prediction head; training the object detection network by taking the initial sample data set as a training sample; finally, detecting inspection pictures by using a the trained object detection network. A light-weight EfficientNet-B0 backbone network feature extraction method obtains more features of objects. Meanwhile, an introduction of multi-scale feature fusion better adapts to small object detection, and a light-weight y_pred regression classification detection head is effectively deployed and realized in Raspberry Pi embedded equipment with tight resources and limited computing power.

Abstract: Iterative prediction systems and methods for the task of action detection process an inputted sequence of video frames to generate an output of both action tubes and respective action labels, wherein the action tubes comprise a sequence of bounding boxes on each video frame. An iterative predictor processes large offsets between the bounding boxes and the ground-truth.

Abstract: The present disclosure provides a method, apparatus and system for identifying target objects. The method includes: clipping out a target image from an acquired image, wherein the target image involves a plurality of target objects to be identified that are stacked; adjusting a height of the target image to a preset height, wherein a height direction of the target image corresponds to a direction in which the plurality of target objects are stacked; extracting a feature map of the adjusted target image; segmenting the feature map in a dimension corresponding to the height direction of the target image to obtain a preset number of segment features; and identifying the target objects based on each of the preset number of segment features.

Abstract: Disclosed are methods, devices, and computer-readable media for operating paired or grouped drone devices. In one embodiment, a method is disclosed comprising capturing a first image by a camera installed on a first drone device; processing the first image using an artificial intelligence (AI) engine embedded in the first drone device, the processing comprising generating a first inference output; transmitting the first inference output to a second drone device; receiving a second inference output from the second drone device, the second inference output associated with a second image captured by the second drone device; and transmitting the first image to a processor based on the first inference output and second interference output.

Abstract: Selection of an evaluation index based upon statistical values of respective evaluation indices is disclosed. In one example, an information processing apparatus comprises an evaluation index unit that determines a plurality of evaluation indices on a basis of imaging data obtained by imaging a subject, wherein the evaluation indices are respectively based upon different wavelength component combinations of the imaging data. An evaluation value unit determines an evaluation value based on a statistical value of each of the evaluation indices, for each of the evaluation indices. A selection unit then determines a selected evaluation index from the evaluation indices on a basis of the evaluation values.

Abstract: An agricultural method includes providing a positive air pressure chamber to prevent outside contaminants from entering the chamber; growing crops in a plurality of cells in the chamber, each cell having multi-grow benches or levels, each cell further having connectors to vertical hoists for vertical movements in the chamber; maintaining pre-set temperature, humidity, carbon dioxide, watering and lighting levels to achieve predetermined plant growth; using motorized transport rails to deliver benches for operations including seeding, harvesting, grow media recovery, and bench wash; dispensing seeds in the cell with a mechanical seeder coupled to the transport rails; growing the crops with computer controlled nutrients, light and air level; and harvesting the crops and delivering the harvested crop at a selected outlet of the chamber.

Abstract: A method of evaluating one or more kernels of an ear of maize using digital imagery that includes acquiring a digital image of the one or more kernels of the ear of maize without the use of spatial reference points, processing the digital image to estimate at least one physical property of the one or more kernels of the ear of maize from the digital image, and evaluating the at least one kernel of maize using the estimate of the at least one physical property of the at least one kernel of maize. The method includes using one or more such digital images to estimate yield on a plant, management zone, field, county and country level.

Abstract: Apparatus and methods for a smart glasses device are provided. The smart glasses device may execute a prediction model on video data captured by the smart glasses device to retrain the prediction model. The prediction model may be retrained in response to a data synchronization between an output of the prediction model and a gesture captured by the smart glasses device.

Abstract: An event-based sensor having a sensing part that produces events asynchronously. The output signal includes information relating to at least some of the produced events. The method comprises: estimating a rate of production of events by the sensing part; while the estimated rate is less than a threshold, transmitting the signal having a variable rate; and when the estimated rate is above the threshold, transmitting the signal including information relating to only some of the produced events, such that a rate of events for which information is included in the signal remains within the threshold.

Abstract: State of the art techniques in the domain of video analysis have limitations in terms of capability to capture spatio-temporal representation. This limitation in turn affects interpretation of video data. The disclosure herein generally relates to video analysis, and, more particularly, to a method and system for video analysis to capture spatio-temporal representation for video reconstruction and analysis. The method presents different architecture variations using three main deep network components: 2D convolution units, 3D convolution units and long short-term memory (LSTM) units for video reconstruction and analysis. These variations are trained for learning the spatio-temporal representation of the videos in order to generate a pre-trained video analysis module. By understanding the advantages and disadvantages of different architectural configurations, a novel architecture is designed for video reconstruction.

Abstract: Disclosed are a video watermark identification method and apparatus, device, and a storage medium. The method includes: each video frame of multiple video frames of a video is divided into multiple image blocks to obtain an image sequence corresponding to the each video frame; multiple image sequences corresponding to the multiple video frames are input to a target detection model to obtain a classification result of each image block, and a feature vector of the video is obtained according to all classification results of the multiple image blocks; and the feature vector of the video is input to a watermark identification model to obtain a watermark identification probability output by the watermark identification model, and the video containing a watermark is determining in a case where the watermark identification probability is greater than or equal to a probability threshold.

Abstract: The present disclosure provides a method and an apparatus for sampling training data and a computer server. The method includes: inputting a video to a target detection model to obtain a detection result for each frame of image; inputting the detection results for all frames of images in the video to a target tracking model, to obtain a tracking result for each frame of image; and for each frame of image in the video: matching the detection result and the tracking result for the frame of image, and when the detection result and the tracking result for the frame of image are inconsistent with each other, determining the frame of image as a sample image to be marked, for which processing by the target detection model is not optimal.

Abstract: An error detection system for notifying maintenance personnel of errors in a bowling lane operation may include a camera, artificial light source (optional), a computer that sends an output signal. The output signal may be a signal to provide notification to maintenance personnel to fix the error or an electrical signal to the pinsetter or ball return unit to manipulate operation of the same to prevent further damage. In this manner, interruptions and delays to bowlers will be minimized because maintenance personnel is immediately notified of any errors and can resolve the errors before the bowler realizes the error.

Abstract: A method for jockey and horse recognition and tracking. The method includes receiving input images or a sequence of images obtained from horse racing videos or video streams; extracting features from the images by computational methods; locating jockey and horse positions of a target horse in the images by the computational methods; deciding to accept or reject the computed jockey and horse positions according to an acceptance function; and producing the final jockey and horse positions and their associated information by an error correction algorithm.

Abstract: A server includes a processor programmed to: acquire first metadata of a first media file recorded by a first mobile device; acquire second metadata of a second media file recorded by a second mobile device; determine that the first media file and the second medial file are likely recordings of the same event when a similarity exceeds a first threshold. The processor is further programmed to, when the first media file and the second medial file are likely recordings of the same event: determine, based on a comparison between the first media file and the second media file, which of the first media file and the second media file is a higher quality recording of the same event; and when the first media file is the higher quality recording, send a link to the first media file to the second mobile device.

Abstract: Various techniques are provided for counting and/or tracking objects within a field of view of an imaging system, while excluding certain objects from the results. A monitoring system may count or track people identified in captured images while utilizing an employee identification system including a wireless signal receiver to identify and remove the employees from the result. The system includes algorithms for separating employee counts from customer counts, thereby offering enhanced tracking analytics.