Abstract: Systems and methods for viewing, storing, transmitting, searching, and editing application-specific audiovisual content (or other unstructured data) are disclosed in which edge devices generate content on the fly from a partial set of instructions rather than merely accessing the content in its final or near-final form. An image processing architecture may include a generative model that may be a deep learning model. The generative model may include a latent space comprising a plurality of latent codes and a trained generator mapping. The trained generator mapping may convert points in the latent space to uncompressed data points, which in the case of audiovisual content may be generated image frames. The generative model may be capable of closely approximating (up to noise or perceptual error) most or all potential data points in the relevant compression application, which in the case of audiovisual content may be source images.

Abstract: Disclosed are systems, apparatuses, and methods for and related to dynamic control of torque and or lateral thrust applied to a load suspended load on a suspension cable to thereby achieve a target orientation or position or to otherwise move with reduced self-induced cyclic motion, with better behavior, and or to respond to changes in the load or disturbance forces.

Abstract: A system and a method for optically detecting a pose of at least one instrument in an operating theater are provided. The method includes determining pose information of the at least one instrument with an optical pose detection device of a surgical microscope or with a microscope-external optical pose detection device, determining whether the pose information is biunique or whether a biunique determination of the pose of the instrument is possible, and evaluating additional information for determining additional pose information, at least for a case where no biunique determination of the pose is possible.

Abstract: A method for encoding attributes of points of a point cloud, includes: determining a true transform coefficient by applying RAHT to attributes of a present node; determining a difference between attributes of a present parent node and each parent node of a first set of parent nodes sharing a face or an edge with the present node; selecting a second set of parent nodes from the first set of parent nodes according to the difference; determining a predicted value for the attributes of the present node according to attributes of the second set of parent nodes; determining a predicted transform coefficient by applying the RAHT to the predicted value of the attributes of the present node; determining a residual from the true transform coefficient and the predicted transform coefficient; and encoding the residual to produce encoded data of attributes of the point cloud for a bitstream.

Abstract: In an image recognition apparatus, a processor performs, based on an input image and using an image recognition model, a plurality of object detection processes to detect as an object detection region a region in the input image where a recognition target object is judged to be present. In the plurality of object detection processes, a plurality of mutually different image recognition models are used. The processor generates inference result data according to the degree of overlap among a plurality of object detection regions detected in the plurality of object detection processes.

Abstract: Provided is a method and device for detecting defect, a computer readable storage medium and an electronic device, the method including: acquiring (S310) a detection task, and acquiring various types of images corresponding to the detection task; acquiring (S320) defect detection models trained by a same initial model corresponding to the types of the images respectively; and obtaining (S330) defect detection results by performing defect detection on respective type of images using the defect detection model corresponding to the type of the images.

Abstract: Systems and methods of PET attenuation correction using low-field MR image data includes receiving a first set of image data and a set of low-field magnetic resonance (MR) image data. An attenuation correction map is generated from the low-field MR image data using a first trained neural network. At least one attenuation correction process is applied to the first set of image data based on the attenuation correction map to generate at least one clinical attenuation-corrected image.

Abstract: Methods, electronic devices and computer program products for model processing are disclosed in embodiments herein. A method in an illustrative embodiment includes encoding first data of a point cloud model to obtain a first matrix, and encoding second data of the point cloud model to obtain a second matrix, where the first data and the second data are data of the point cloud model acquired from different angles. The method further includes respectively decomposing, by an equivariant encoder, the first matrix and the second matrix into a first equivariant matrix and a second equivariant matrix, and respectively decomposing, by an invariant encoder, the first matrix and the second matrix into a first invariant matrix and a second invariant matrix. The method further includes training the equivariant encoder and the invariant encoder based on the first equivariant matrix, the second equivariant matrix, the first invariant matrix, and the second invariant matrix.

Abstract: A method and system for communicating estimated blood loss parameters of a patient to a user, the method comprising: receiving data representative of an image, of a fluid receiver; automatically detecting a region within the image associated with a volume of fluid received at the fluid receiver, the volume of fluid including a blood component; calculating an estimated amount of the blood component present in the volume of fluid based upon a color parameter represented in the region, and determining a bias error associated with the estimated amount of the blood component; updating an analysis of an aggregate amount of the blood component and an aggregate bias error associated with blood loss of the patient, based upon the estimated amount of the blood component and the bias error; and providing information from the analysis of the aggregate amount of the blood component and the aggregate bias error, to the user.

Abstract: A booster engine enhances the quality of a frame sequence. The booster engine receives, from a first stage circuit, the frame sequence with quality degradation in at least a frame. The quality degradation includes at least one of uneven resolution and uneven frame per second (FPS). The booster engine queries an information repository for reference information on the frame, using a query input based on at least a region of the frame to obtain a query output. The booster engine then applies a neural network to the query input and the query output to generate an optimized frame, and sends an enhanced frame sequence including the optimized frame to a second stage circuit.

Abstract: The driving assistance system includes an imaging device capable of capturing a first monochrome image in a vehicle traveling direction, a first neural network for segmentation processing, a second neural network for depth estimation processing, a determination portion determining a center of a portion cut off from the first monochrome image on the basis of the segmentation processing and the depth estimation processing, a third neural network for colorization processing of only a second cut-off monochrome image, and a display device for enlargement of the second monochrome image subjected to the colorization processing.

Abstract: There is provided an information processing apparatus that includes a preprocessing unit that determines whether or not to permit provision of a service based on emotion recognition on the basis of a predetermined criterion, an emotion estimation processing unit that performs the emotion recognition for a user by using an artificial intelligence function, and a service providing processing unit that provides a service based on an emotion recognition result by the emotion estimation processing unit, in which, when the preprocessing unit determines to permit the provision of the service, the emotion estimation processing unit performs the emotion recognition or the service providing processing unit provides the service.

Abstract: A three-dimensional data encoding method includes: encoding information of a target node included in an N-ary tree structure of three-dimensional points included in three-dimensional data; and generating a bitstream including the information of the target node encoded. In the encoding, the target node is encoded based on reference limitation information indicating a referable neighbor node among neighbor nodes spatially neighboring the target node. In the generating, when the target node is encoded by reference to information of a first neighbor node, the bitstream further including encoding processing information is generated, the encoding processing information indicating a processing method in the encoding; and when the target node is encoded without reference to the information of the first neighbor node, the bitstream is generated without including the encoding processing information in the bitstream.

Abstract: The quality of a frame sequence is enhanced by a booster engine collaborating with a first stage circuit. The first stage circuit adjusts the quality degradation of the frame sequence when a condition in constrained resources is detected. The quality degradation includes at least one of uneven resolution and uneven frame per second (FPS). The booster engine receives the frame sequence from the first stage circuit, and generates an enhanced frame sequence based on the frame sequence for transmission to a second stage circuit.

Abstract: An annunciation method includes the steps of obtaining first position information as position information of a place where a first object is located, obtaining second position information as position information of a place where a second object is located, obtaining third position information as position information of a place where a third object is located, setting the first object and the second object as a first group when a distance between the first object and the second object calculated based on the first position information and the second position information is smaller than or equal to a first threshold value, obtaining fourth position information representing a position of the first group, and performing annunciation when a distance between the third object and the first group calculated based on the third position information and the fourth position information is smaller than a second threshold value.

Abstract: Rendering an avatar may include determining an expression to be represented by an avatar, obtaining a blood texture map associated with the expression, wherein the blood texture map represents an offset of coloration from an albedo map for the expression, and rendering the avatar utilizing the blood texture map.

Abstract: Various implementations track an eye using both camera images and light sensor data, e.g., from a photosensitive surface or set of photodetectors. The camera need not capture images at a high capture rate since the light sensor data captured from the light sensor provides information about the eye during the time periods between camera frames. An eye tracking device may thus provide eye tracking with high temporal resolution using less power and resources and prior devices. The eye tracking device includes a lens (e.g., a diffractive optical element (DOE)) that splits received light onto the camera and light sensor. The camera and light sensor may be aligned, e.g., co-centered. Using such a lens to provide light to an aligned camera and light sensor facilitates tracking the eye characteristics by reducing the need for matching the data and/or accounting for multiple environmental light sources.

Abstract: A three-dimensional data encoding method includes: assigning three-dimensional points to one of layers, based on items of geometry information of the three-dimensional points; searching three-dimensional points surrounding a current three-dimensional point to be encoded, to select, from the three-dimensional points, a three-dimensional point to be referred to when a predicted value of attribute information of the current three-dimensional point is calculated, the current three-dimensional point belonging to a first layer among the layers; and calculating the predicted value of the attribute information of the current three-dimensional point using the three-dimensional point selected. In the searching of the three-dimensional points, a search range for a same layer as the current three-dimensional point is different from a search range for a layer higher than the first layer.

Abstract: According to an embodiment, a method for generating a digital data set for fabricating a physical dental bleaching tray useable to deliver an bleaching agent is disclosed. The method includes obtaining a three-dimensional digital representation of a patient's dentition including teeth and gingiva; segmenting two or more teeth into individual tooth; identifying a facial surface of at least one of the segmented tooth; defining a facial surface portion including a surface area that is at least partly bound by a virtual boundary that is non-interfacing with the gingiva; generating a modified three-dimensional digital representation using the defined facial surface portion; and generating, based on the modified three-dimensional digital representation, the digital data set configured to be used in fabricating the physical dental bleaching tray.

Abstract: Items are identified in a waste stream for purposes of recycling, using deterministic and/or probabilistic techniques. Imagery of the waste stream from multiple viewpoints permit creation of a 3D depth draped image representation, from which one or more 2D planes can be synthesized. Phase-coherent patches of recoverable encoded data can be identified from among soiled and crumpled object surfaces, and used in combination to recover object identification information. Recognition of certain items can trigger further image processing that is specific to such items. (Detection of a catsup bottle, for example, can trigger image analysis to discern the presence of catsup residue.) Information about recognized objects can be provided to external data customers, e.g., to track grey market diversion of particular products into unlicensed territories. These and other features and advantages, which can be used alone or in combination, are detailed herein.