Abstract: A method for generating a semantically modified variation of an image. In the method: the image is divided into equally sized, non-overlapping patches; the patches are converted with a patch encoding function of a transformer network into a chain of tokens in a workspace; the tokens are grouped into preservation tokens, whose information is to be preserved in the variation, and masked tokens, whose information is to be masked in the variation; the preservation tokens are converted with an encoder of the transformer network into a chain of processed tokens; the chain is supplemented through application of an insertion operator that inserts these masked tokens at positions corresponding to the positions of the masked tokens in the original chain to form a chain that represents the sought variation; the chain is converted with a decoder of the transformer network into the sought variation.

Abstract: Methods and systems for generating orthographic imagery are provided. An example method involves accessing a plurality of source images depicting a scene from multiple points of view, encoding at least some of the source images into corresponding series of multiscale feature maps, receiving a definition of an orthographic projection of the scene, and generating an orthographic image of the scene corresponding to the orthographic projection. The generating involves applying global attention across a set of higher-level features, applying a convolutional and upsampling layer, generating a depth map for the scene, and back-projecting each point on the depth map corresponding to a pixel that should be rendered back through the orthographic projection to determine a set of lower-level features to be used to decode the pixel, and applying local attention across the set of lower-level features to produce a set of decoded features from which pixel information can be determined.

Abstract: Information analysis and display systems and methods include receiving first financial information regarding a user, and calculating a target home price based on the received financial information. A user interface display is generated that displays the target home price, and modifications to the target home price from the user are received via the user interface. Homes for sale are evaluated based on the target price.

Abstract: A computer-implemented method of changing a face within an output image or video frame includes: receiving an input image that includes a face presenting a facial expression in a pose; separately encoding different portions of the image by, for each separately encoded portion, generating a latent space point of the portion, thereby generating a plurality of multi-dimensional vectors where each multi-dimensional vector is an encoded representation of a different portion of the input image; concatenating the plurality of multi-dimensional vectors into a combined latent space vector; and decoding the combined latent space vector to generate the output image in accordance with a desired facial identity but with the facial expression and pose of the face in the input image.

Abstract: Training an ordinal mapping deep neural network (OMDNN) can include receiving multiple samples, each a computer-processable data structure corresponding to a real-world object and including a data element indicating one of n predefined classes to which each sample is linked. Each sample can be mapped by the OMDNN to sample points of a multidimensional space. The OMDNN can predicts the class of each sample based on an ordinal mapping. Parameters of the OMDNN can be iteratively adjusted in response to misclassifying one or more samples. Iteratively adjusting the parameters can be based on an expected loss determined by an ordinal mapping loss function that measures (a) distances between each sample point in the multidimensional space and each other sample point of the same class and (b) overlap between sample points of different classes.

Abstract: A device, method, system, or article has multi-directional rolling cache to store image data.

Abstract: A data processing apparatus includes a control unit. The control unit is configured to obtain capture image data generated based on image capturing by a camera provided for a robot, combine at least a portion of display image data of the capture image data obtained at a first time point and at least a portion of past capture image data of the capture image data obtained at a second time point before the first time point to generate composite image data having a wider angle of view than the display image data, and display the composite image data on a display device viewable by an operator of the robot.

Abstract: A computer-implemented method of determining whether digital content has been viewed by a subject, the method including, displaying, by a display device, the digital content to the subject; tracking the gaze of the subject whilst displaying at least a temporal portion of the digital content to acquire tracking data; determining from the tracking data, the display portions of the display device viewed in that temporal portion of the digital content; and generating at least one score by comparing these portions to a predetermined template of display portions for that temporal portion of the digital content; and determining that the subject has viewed at least the temporal portion of the digital content based on the at least one score. A display device and system are also disclosed.

Abstract: This disclosure relates to the combining and interaction of multiple artificial intelligence (AI) models for medical image analysis. An example method includes obtaining AI models from model providers and organizing them to form associations. In response to a user request, base models are selected and provided. Additional models are further selected to combine with the base models, and medical image analysis results are presented based on applying a combination of the models to target medical image data.

Abstract: Embodiments of the present disclosure relate to sequentially loading keypoint descriptors of a previous image and comparing them with a plurality of keypoint descriptors of a current image. The keypoint descriptors of the previous image are stored and accessed from a system memory while the keypoint descriptors of the current image are stored and accessed from a local memory. Hence, the keypoint descriptors of the previous image are received only once at a descriptor match circuit while the keypoint descriptors of the current image are received multiple times for comparison against different keypoint descriptors of the previous image.

Abstract: According to an aspect of the disclosure, there is provide a computer implemented method including: receiving via a UI screen a first instruction to set a plurality of layer images and a response amount for a visual effect, the plurality of layer images being to be included in a composite image, the visual effect corresponding to an amount of a user operation specified by a user action for each of the plurality of layer images included in the composite image, the visual effect being applied regardless of selection of any one of the plurality of layer images; controlling a display device to display the composite image in which the plurality of layer images are stacked, the visual effect of each layer image being changed in response to the user action; and receiving a second instruction to store data, the data including the plurality of layer images or information indicating a location of the plurality of images and data indicating the response amount.

Abstract: This application relates to the field of photographing technologies, and discloses an image processing method and an electronic device, so as to dynamically adjust LUTs in a process of photographing or visual recording, and enrich display effects of photographing or visual recording. The electronic device obtains a first image, where the first image is an image captured by a camera of the electronic device, and the first image includes a first shot object; the electronic device determines a first scenario corresponding to the first image, where the first scenario is used to identify a scenario corresponding to the first shot object; the electronic device determines a first LUT based on the first scenario; and the electronic device processes the first image based on the first LUT to obtain a second image, and displays the second image, where a display effect of the second image corresponds to the first LUT.

Abstract: Various embodiments of the present disclosure relate generally to systems and methods for generating multi-view interactive digital media representations in a virtual reality environment. According to particular embodiments, a plurality of images is fused into a first content model and a first context model, both of which include multi-view interactive digital media representations of objects. Next, a virtual reality environment is generated using the first content model and the first context model. The virtual reality environment includes a first layer and a second layer. The user can navigate through and within the virtual reality environment to switch between multiple viewpoints of the content model via corresponding physical movements. The first layer includes the first content model and the second layer includes a second content model and wherein selection of the first layer provides access to the second layer with the second content model.

Abstract: A method for transparentizing a target object of a video includes following steps performed by a processor: extracting a target block corresponding to an target object from a frame of the video; setting a target transparentizing weight corresponding to the target block; performing a dilation operation upon the target block to generate a dilation outline enclosing the target block; setting a dilation transparentizing weight according to a dilation region, wherein the dilation region is a region inside the dilation outline and outside the target block; cutting a background block from the frame according to the dilation outline; and performing an overlapping operation according to a previous background block, the background block, the target block, the target transparentizing weight, the dilation region, and the dilation transparentizing weight to generate an output frame, wherein the previous background block does not correspond to the target object.

Abstract: The present application describes machines and methods that leverage enabling technologies such as robotics, sensing, machine learning, and wireless internet coverage (e.g., 5G cell coverage) in order to monitor photovoltaic systems such as solar sites. Machines according to the present disclosure can be operated remotely by users to traverse a solar site and perform a series of inspection steps, such as via an online portal. Autonomous and semi-autonomous operations are also possible. These methods and machines can eliminate the need for a technician to visit the site in person for routine inspection, and can provide better information when a site alarm is triggered so that if a technician does need to visit the site, he or she is better prepared. Also described herein are systems and methods for inventorying solar sites, and systems and methods for inspection of structure-mounted photovoltaic systems.

Abstract: The invention provides a method for generating multilayer images. The method comprises receiving three-dimensional, 3D, source data of a scene and generating a plurality of layers at different depths in a virtual 3D scene, wherein each layer corresponds to a particular layer depth value relative to an origin point. Image data is generated for each layer based on the 3D source data, wherein the image data comprises texture data and transparency data, and displacement maps are generated for the layers based on the 3D source data and the depth values of the layers. The displacement maps define the depth of different points on a corresponding layer relative to the layer depth, wherein the generation of the displacement maps is restricted such that a depth surface, as defined by any displacement map, does not intersect any other depth surface in the virtual 3D scene. A depth surface is defined as the mapping of a displacement map onto the corresponding layer.

Abstract: An automated cage-to-cage fitting technique is used to fit arbitrary three-dimensional (3D) geometry to arbitrary target 3D geometry for 3D avatars in a 3D environment, such as an online game. Each graphical representation of items of clothing is provided with an inner cage and with an outer cage. The inner cage of a clothing item (a current layer) is mapped to the outer cage of another clothing item (a previous/underlying layer) that has already been fitted on the 3D avatar. This mapping enables the current layer to conform to the shape of the previous layer. The technique allows any body geometry to be fitted with any clothing geometry, including enabling layers of clothing to be fitted over underlying layer(s) of clothing, thereby providing customization.

Abstract: An autostereoscopic display apparatus includes an array of display pixels arranged in rows and columns; an array of elongated optical elements extending parallel to one another and overlying the display pixels, wherein the optical elements are slanted at an angle relative to the columns of display pixels; for any given display pixel, an optimized brightness is determined based on the distance between the pixel and the nearest end of a boundary separating adjacent images. By adjusting pixel brightness, the apparatus effectively reduces image crosstalk, thereby enhancing the clarity and quality of autostereoscopic displays.

Abstract: Various implementations disclosed herein include devices, systems, and methods that render perspective correct 2D graphics (e.g., virtual objects, text, graphics) correctly at any angle and distance in an electronic device such as an HMD. In some implementations, a vector graphic including portions defined on a 2D canvas is obtained and the 2D canvas is divided into separate regions, each region containing a plurality of pixels. In some implementations, data based on associating a plurality of the regions of the 2D canvas with the portions of the vector graphic is generated, the data identifying one or more contributing portions of the portions of the vector graphic to use in rendering the pixels in each of the plurality of regions of the 2D canvas. In some implementations, the data is provided to a processor for rendering the vector graphic.

Abstract: A system and method of automatic image view alignment for a camera-based road condition detection on a vehicle. The method includes transforming a fisheye image into a non-distorted subject image, comparing the subject image with a reference image, aligning the subject image with the reference image, and analyzing the aligned subject image to detect and identify road conditions in real-time as the vehicle is in operation. The subject image is aligned with the reference image by determining a distance (d) between predetermined feature points of the subject and reference images, estimating a pitch of a projection center based on the distance d, and generating an aligned subject image by applying a rectification transformation on the fisheye image by relocating a center of projection of the fisheye image by the pitch angle .

Abstract: Systems and methods are provided where requests to modify or validate a road network map are received, and a determination is made as to whether a requested modification, or the road network map, is invalid based on a location at which the modification is desired to be made, and rules associated with a corresponding bounded area of the road network map to which the location belongs. Where a requested modification, or the road network map, is determined to be invalid, the modification is inhibited, or a notification is displayed to a user that the road network map violates the set of rules.

Abstract: A display system can be configured to receive, via a user interface, a first identifier associated with a first space object and determine a first maneuver of the first space object. The first maneuver can include a perturbation of the path of the first space object. Based on the first identifier and the first maneuver, the display system can identify one or more path parameters associated with a path of the first space object and generate a display interface. The display interface can include a longitude-time graph having a longitude axis spanning from a lower-longitude limit to an upper-longitude limit and a time axis spanning from the lower-time limit to the upper-time limit and an indication of the one or more path parameters.

Abstract: In one embodiment, a method includes presenting, to a user in an XR environment, a persistent GUI element that includes a control deck. The control deck includes one or more first interactive elements, each first interactive element associated with a particular function. The method further includes locking, in a first control-deck mode, the control deck to the user such that the control deck appears at the same relative position to the user regardless of the user's movements in the XR environment.

Abstract: A method and system for generating a random graphic training dataset and a hologram is disclosed, which relates to the technical field of image processing. The method for generating a random graphic training dataset includes: randomly generating a plurality of graphic datasets, where each of the graphic datasets includes a square, a circle and an equilateral triangle; for any graphic dataset, performing gradual filling on each graphic in the graphic dataset to obtain a gradual graphic dataset; and performing translation transformation on each graphic in the gradual graphic dataset for a predetermined number of times to obtain random graphics corresponding to the graphic dataset, and determining random graphics corresponding to all graphic datasets as a random graphic training dataset. More comprehensive training data can be provided and a training effect of a model can be improved.

Abstract: A technique for performing style transfer between a content sample and a style sample is disclosed. The technique includes applying one or more neural network layers to a first latent representation of the style sample to generate one or more convolutional kernels. The technique also includes generating convolutional output by convolving a second latent representation of the content sample with the one or more convolutional kernels. The technique further includes applying one or more decoder layers to the convolutional output to produce a style transfer result that comprises one or more content-based attributes of the content sample and one or more style-based attributes of the style sample.

Abstract: Systems and methods described herein provide for retrieving, from a storage device, first image data previously captured by a client device. The systems and methods further detect a selection of a first image processing operation and perform the first image processing operation on the first image data to generate second image data. The systems and methods further detect a selection of a second image processing operation and perform the second image processing operation on the second image data to generate third image data. The systems and methods generate a message comprising the third image data.

Abstract: A vehicle may include a movable roof, a sensor supported by the roof, and an actuator for selectively raising and lowering the roof.

Abstract: A method includes obtaining first data representing a subterranean domain, and creating a geotag associated with a three-dimensional location of interest in the subterranean domain and a feature at the location of interest. The geotag is further associated with second data that describes the location of interest in the subterranean domain. The method also includes generating metadata describing a portion of the first data, the second data, or both, storing the geotag in a database including a plurality of geotags. The method may also include searching for an analog to the geotag.

Abstract: A vehicle display device is configured to superpose a background layer on a first layer for displaying map and/or periphery information, and superpose a layer for displaying information other than the map information and the vehicle periphery information on the background layer. In accordance with a user operation, the display is switched from a transparent background layer where the user can see the information displayed on the first layer to an opaque background layer where the user cannot see the information displayed on the first layer. While switching between the transparent background layer to the opaque background layer, a visual effect is not added to show the switching. After the switch, a transmittance of the background layer is changed and a visual effect is not added to show the switching of the background layer.

Abstract: Medical software tools platform utilizes a surgical display to provide access to specific medical software tools, such as medically-oriented applications or widgets, that can assist those in the operating room, such as a surgeon or surgical team, with a surgery. For various embodiments, the medical software tools platform and its associated medical software tools are presented on a surgical display (e.g., being utilized in an operating room) over an image stream provided by a surgical camera (e.g., in use in the operating room or with other endoscopic procedures). Various medical software tools can provide features and functions that can facilitate integration of equipment in an operating room or add medical context awareness to anatomic structures presented in the image stream from the surgical camera and provide archived information pertaining to the same.

Abstract: An inspection system for a welded portion in welding includes a display unit, an acquisition unit that acquires a reference position of a welded portion stored in the storage device, a display control unit that displays, on the display unit, a reference position image that is an image showing a reference position virtually superimposed on an image of an object to be welded in a real space captured by the imaging device and includes a welded portion, and a distance calculation unit that calculates a distance in the real space between the welded portion in the real space displayed on the display unit and the reference position of the welded portion.

Abstract: Technologies for transitioning between two-dimensional (2D) and three-dimensional (3D) display views for video conferencing are described herein. Video conferencing applications can have multiple display views for a user participating in a video conference. In certain situations a user may want to transition from a 2D display view of the video conference to a more immersive 3D display view. These transitions can be visually jarring and create an uncomfortable user experience. The transition from a 2D display view to a 3D display view can be improved by executing the transition to a 3D display view by manipulating visual properties of a virtual camera that is employed to generate the display views.

Abstract: Methods and systems for removing sensitive information from a digital image. An instruction to share a digital image is received. It is then determined that the digital image contains a depiction of a corporate display medium that is classified as sensitive based on a policy and, based on the determination that the digital image contains the depiction of the corporate display medium that is classified as sensitive based on the policy, the digital image is processed to modify the depiction. The digital image is shared.

Abstract: A semiconductor device having favorable display quality is provided. The semiconductor device is provided with a display portion, a line-of-sight sensor portion, a control portion, and an arithmetic portion. The line-of-sight sensor portion has a function of obtaining first information showing a direction of a user's line of sight. The arithmetic portion has a function of determining a first region including a gaze point of the user on the display portion with use of the first information and a function of increasing a definition of an image displayed on the first region. Light emitted from the display portion may be used to obtain the first information showing the direction of the line of sight.

Abstract: The present disclosure provides a display in application of a vehicle. The display includes two display devices, in which at least one of the two display devices includes a display panel and a light controller. The display panel includes a plurality of signal lines respectively extending along a first direction. The light controller is disposed adjacent to the display panel and includes a plurality of patterns respectively extending along a second direction, in which the second direction is different from the first direction. The display devices are arranged along the first direction.

Abstract: A server system is communicable with an image forming apparatus and an information processing apparatus, and includes a first reception unit configured to receive image data generated by a first application, a transmission unit configured to transmit, to the information processing apparatus, information to activate a second application that enables a user to perform print setting used in printing of the image data received by the first reception unit, a second reception unit configured to receive the print setting performed in the second application, and a print control unit configured to cause the image forming apparatus to perform printing based on at least the image data received by the first reception unit and the print setting received by the second reception unit.

Abstract: In one example aspect, a method of video processing includes performing a conversion between a current block of a video and a bitstream of the video using a palette mode in which a palette of representative sample values is used for coding the current block. During the conversion, one or multiple predictor palettes are used for predicting the palette of representative sample values, and wherein updating of the one or multiple predictor palettes after the conversion of the current block is disabled according to a rule based on a characteristic of the current block.

Abstract: An example computing platform is configured to (i) receive a two-dimensional (2D) image file comprising a construction drawing, (ii) generate, via semantic segmentation, a first set of polygons corresponding to respective areas of the 2D image file, (iii) generate, via instance segmentation, a second set of polygons corresponding to respective areas of the 2D image file, (iv) generate, via unsupervised image processing, a third set of polygons corresponding to respective areas of the 2D image file, (v) based on (a) overlap between polygons in the first, second, and third sets of polygons and (b) respective confidence scores for each of the overlapping polygons, determine a set of merged polygons corresponding to respective areas of the 2D image file, and (vi) cause a client station to display a visual representation of the 2D image file where each merged polygon is overlaid as a respective selectable region of the 2D image file.

Abstract: A Free Viewpoint Video (FVV) generation and interaction method based on a deep Convolutional Neural Network (CNN) includes the steps of: acquiring multi-viewpoint data of a target scene by a synchronous shooting system with a multi-camera array arranged accordingly to obtain groups of synchronous video frame sequences from a plurality of viewpoints, and rectifying baselines of the sequences at pixel level in batches; extracting, by encoding and decoding network structures, features of each group of viewpoint images input into a designed and trained deep CNN model, to obtain deep feature information of the scene, and combining the information with the input images to generate a virtual viewpoint image between each group of adjacent physical viewpoints at every moment; and synthesizing all viewpoints into frames of the FVV based on time and spatial position of viewpoints by stitching matrices. The method does not require camera rectification and depth image calculation.

Abstract: Centralized management of a Data Processing Unit (DPU) Baseboard Management Controller (BMC) through an integrated server remote access controller (iRAC) may include embedding a secure token in a communication from the iRAC to a BMC of a DPU, the secure token authorizing the iRAC to the DPU BMC and authorizing the DPU BMC to the iRAC. The secure token may include a first layer token authorizing the iRAC to the DPU and authorizing the DPU to the iRAC and a second layer token authorizing the DPU to the DPU BMC and authorizing the BMC to the DPU. Alternatively, the secure token may be generated by the iRAC generating an initial token authorizing the iRAC to the DPU and authorizing the DPU to the iRAC, the iRAC embedding the initial token in a request to the DPU for a resource of the DPU BMC and the DPU generating the secure token.

Abstract: Some implementations include methods generating visualization emphasis object for players and may include receiving a video clip associated with a sporting event participated by a plurality of players using a playing field with a play object. The players in a frame of the video clip may be detected. Players who are on the playing field may be identified from the detected players. Each of the players identified to be on the playing field may be associated with a rectangular bounding box that provides an outline of each of the players. A player who has possession of the play object may be identified. A visualization emphasis object may be generated and placed on the player who has possession of the play object. The visual emphasis object may have a size proportional to height of a bounding box associated with the player having the possession of the play object.

Abstract: A third-party user input content item is presented on a user device of a user. A first content augmentation of the third-party user input content item by a first content augmentation function is determined. A camera graphical element is presented. The camera graphical element is user-selectable to invoke a camera function and cause presentation of a first content augmentation selection graphical element. Responsive to determining a user selection of the camera graphical element, the camera function is invoked. The user is enabled to select a second user input content item and the first content augmentation selection graphical element is user-selectable to invoke the first content augmentation function.

Abstract: The disclosure discloses a method, a medium, and a device for enhancing event camera image reconstruction by fusing visible images, and belongs to the field of image processing. In order to solve the problem that the reconstructed image of the event camera is likely affected by noise and lacks of photorealistic, the method provided by the disclosure adopts an enhancement module based on the improved Transformer model to fuse the intensity image reconstructed from the event stream and the reference image selected from the visible image sequence captured by an optical sensor, so that makes the reconstructed image has an appearance close to the visible image while maintaining the characteristics of high dynamic range of the event camera.

Abstract: One embodiment provides a method comprising, at a runtime library executed by a processor of a data processing system, receiving an input frame having objects to be stylized via a style transfer network associated with the runtime library, wherein the style transfer network is a neural network model trained to apply one or more visual styles to an input frame, performing instance segmentation on the input frame to generate one or more instance masks to identify one or more objects to be stylized, generating one or more stylized frames for each style to transfer to the input frame, and merging, via the one or more instance masks, stylized objects from one or more stylized frames with un-stylized content from the input frame to generate an output frame with per-instance stylization.

Abstract: A method is performed at an electronic device with a one or more processors, a non-transitory memory, and a display. The method includes presenting, on the display, a plurality of objects including a first object and a second object. The method includes obtaining a first display priority value that is associated with the first object. The method includes prioritizing the first object over the second object based on a function of the first display priority value. The method includes, in response to determining that each of the first object and the second object satisfies an offscreen criterion, displaying, on the display, a first offscreen indicator that is associated with the first object according to the prioritization.

Abstract: A virtual, augmented, or mixed reality display system includes a display configured to display virtual, augmented, or mixed reality image data, the display including one or more optical components which introduce optical distortions or aberrations to the image data. The system also includes a display controller configured to provide the image data to the display. The display controller includes memory for storing optical distortion correction information, and one or more processing elements to at least partially correct the image data for the optical distortions or aberrations using the optical distortion correction information.

Abstract: Methods and systems for enlarging a stylized region of an image are disclosed that include receiving an input image, generating, using a first generative adversarial network (GAN) generator, a first stylized image, based on the input image, normalizing the input image, generating, using a second generative adversarial network (GAN) generator, a second stylized image, based on the normalized input image, blending the first stylized image and the second stylized image to obtain a third stylized image, and providing the third stylized image as an output.

Abstract: Provided are a display control apparatus, a display control method, and a program that enable developers to implement application programs without needing to be aware of a safe area. An image acquisition section (42) acquires an image that is drawn in a frame buffer by an application program. A safe area setting specification section (44) specifies the setting of a safe area in a display section. An image generation section (48) generates a display target image including, as a part, the image sized according to the setting of the safe area. A display control section (50) causes the display section to display the display target image.

Abstract: A compact generative neural network can be distilled from a teacher generative neural network using a training network. The compact network can be trained on the input data and output data of the teacher network. The training network train the student network using a discrimination layer and one or more types of losses, such as perception loss and adversarial loss.

Abstract: An embodiment user terminal includes an image acquisition part, a user interface configured to display an image photographed through the image acquisition part, a position detection sensor configured to detect a position of the user terminal, and a controller configured to determine a recommended point of interest (POI) among a plurality of POIs located around the user terminal, based on operation information about each of the plurality of POIs, a user's search history for a POI, and the user's POI scrap, and to control the user interface to display an augmented reality (AR) image corresponding to the recommended POI by superimposing the AR image on the image photographed through the image acquisition part.