Abstract: It is desirable to provide a technology making it possible to enhance noticeability of a virtual object. Provided is an information processing apparatus that includes a display controlling section that causes a display device to start displaying a first virtual object while a user is moving in a real space, performs distance keeping control to substantially keep a distance between the user and the first virtual object, ends the distance keeping control in a case where a predetermined condition is satisfied, and changes the distance between the user and the first virtual object.

Abstract: A computer implemented method is disclosed for identifying a pose of a probe by registering an ultrasound image with volumetric scan data. The volumetric scan data is processed (31) to determine simulated ultrasound images corresponding with a plurality of different poses of the probe. A feature vector is extracted (32) from each of the simulated ultrasound images and from the ultrasound image. The feature vector from the ultrasound image is compared (33) with each feature vector of the simulated ultrasound images to determine a distance or similarity for each simulated ultrasound image. At least one candidate image is selected (34), the at least one candidate image comprising a subset of the simulated ultrasound images that best matches the ultrasound image, based on the distance or similarity. The pose of the probe is identified (35) from the at least one candidate image.

Abstract: Disclosed is a system and method for presenting a graphical representation of an oral situation of a patient over time. In particular the system and method relates to a method of presenting a plurality of scans taken over time in manner where the focus on the changes in the oral situation of the patient is maintained.

Abstract: Provided are a scrap determination system and method that can improve scrap determination techniques. A scrap determination system (1) comprises: a first scrap determination model (221) generated using teaching data including first learning images, and determines, based on a camera image, grade of scrap in the image and a ratio of the grade; a second scrap determination model (222) generated using teaching data including second learning images, and determines, based on the image, grade of scrap in the image and a ratio of the grade; a selection model (223) configured to determine which of the first scrap determination model (221) and the second scrap determination model (222) is to be used, based on the image; and an output section (24) configured to output information of each grade of scrap and a ratio of the grade determined based on the image using the model selected by the selection model (223).

Abstract: Systems and techniques are provided for deep learning model optimizations. An example process can include generating, based on processing data using a neural network model, a plurality of prediction outputs associated with a plurality of output channels corresponding to a multi-channel prediction target; determining that a prediction output from the plurality of prediction outputs has a value that is outside of a quantization range and one or more remaining prediction outputs from the plurality of prediction outputs have a respective value that is within the quantization range; clamping the prediction output based on the quantization range; and generating a single channel output based on the clamped prediction output and the one or more remaining prediction outputs.

Abstract: Training of a machine learning model is included. An electronic device includes a memory storing a machine learning model including a discriminator model, and a processor configured to extract, from a training image, a first image patch and a second image patch at least partially overlapping the first image patch, extract, from the first image patch, a first feature map based on a layer of the discriminator model, extract, from the second image patch, a second feature map based on the layer, extract a first partial feature map from a projected map that is projected based on the first feature map, and train the discriminator model based on a first objective function value generated based on a second partial feature map and the first partial feature map, wherein the second partial feature map corresponds to a part of the extracted second feature map.

Abstract: A process for right-sizing a software stack includes: identifying a first wireless communications node (1WCN) for a wireless communications system (WCS) that includes a second wireless communications node (2WCN), obtaining a first adaptive software stack (1ASS) which configures the Device to communicate data to a Destination using the 1WCN, and determining if the 1ASS optimizes the Device for the communicating of the data with the Destination via the 1WCN. If not optimized, a second adaptive software stack (2ASS) may be obtained and a determination made as to whether the 2ASS second optimizes the Device. At least one of the 1ASS and the 2ASS may be implemented and the Device is coupled with the 1WCN by a communications links. A first application may be selected and executed. During software execution application data may be communicated with the Destination using the 1WCN and the as implemented 1ASS or 2ASS.

Abstract: An image processing device includes an input/output interface, and at least one processor. The at least one processor executes: detecting feature points in a facial image of an object included in an image taken from the input/output interface; acquiring first distance information between the feature points before a beauty treatment is performed; acquiring second distance information between the feature points at a second timing after the beauty treatment is performed; acquiring a difference value between the first distance information and the second distance information; and determining whether the beauty treatment is correctly given or has been given based on a different value between the acquired distance information.

Abstract: A base station antenna includes a first antenna having first and second spaced-apart columns of first radiating elements therein, which are configured to operate within a first frequency band. An active antenna system (AAS) is provided, which is configured to operate within a second, typically higher, frequency band. The AAS includes a second antenna within a space between the first and second columns of first radiating elements. These first radiating elements may include tilted feed stalks which support higher integration by enabling the first radiating elements to overhang at least a portion of the second antenna.

Abstract: Methods and systems for providing a dataset of human in-bed poses include simultaneously gathered images of in-bed poses of humans from imaging modalities including red-green-blue (RGB) and one or more of long wavelength infrared (LWIR), depth imaging, and pressure mapping. The images are obtained under a lighting condition and a cover condition. The dataset can be used to train a model of estimating human in-bed poses and for methods of estimating human in-bed poses. Methods and systems of estimating three-dimensional human poses from two-dimensional input images are provided.

Abstract: A computer-implemented method of reducing noise in magnetic resonance (MR) images is provided. The method includes executing a neural network model of analyzing MR images, wherein the neural network model is trained with a pair of pristine images and corrupted images. The pristine images are the corrupted images with noise reduced, and target output images of the neural network model are the pristine images. The method also includes receiving first MR signals and second MR signals, reconstructing first and second MR images based on the first MR signals and the second MR signals, and analyzing the first MR image and the second MR image using the neural network model. The method further includes deriving a denoised MR image based on the analysis, wherein the denoised MR image is a combined image based on the first MR image and the second MR image and outputting the denoised MR image.

Abstract: The present disclosure provides methods and apparatuses for image processing and image classification. In one embodiment, the method for image processing comprises: receiving an image; obtaining a first classification result for the image based on a classification model; processing the image for classification based on a preset process, and providing a processing result into a re-ranking model to obtain a second classification result for the image; and determining a target classification result for the image, based on the first classification result and the second classification result.

Abstract: Methods, apparatus, systems, and computer-readable media are provided for generating and/or adapting automated assistant content according to a distance of a user relative to an automated assistant interface that renders the automated assistant content. For instance, the automated assistant can provide data for a client device to render. The client device can request additional data when the user relocates closer to, or further from, the client device. In some implementations, a request for additional data can identify a distance between the user and the client device. In this way, the additional data can be generated or selected according to the distance in the request. Other implementations can allow an automated assistant to determine an active user from a group of users in an environment, and determine a distance between the active user and the client device in order that any rendered content can be tailored for the active user.

Abstract: A method for visualization of digitized slides is disclosed. The method comprises retrieving, by one or more computer processors, a digitized slide, determining, by the one or more computer processors, one or more visualization components of the digitized slide, generating, by the one or more computer processors, a virtual slide corresponding to the digitized slide based on the one or more visualization components, and displaying, by the one or more computer processors, a visualization of the virtual slide.

Abstract: A spinal surgery training process includes the steps of capturing a plurality of 2D images for each of a plurality of spines, generating a curve of each spine from the respective 2D images based on locations of select vertebrae in each of the spines, grouping the spines into one of a number of groups based on similarity to produce groups of spines having similarities, performing the capturing, generating, determining and grouping steps at least once prior to surgery and at least once after surgery to produce pre-operative groups and their resultant post-operative groups, and assigning surgical methods and a probability to each of the post-operative groups indicating the probability that a spinal shape of the post-operative group can be achieved using the surgical methods. An outcome prediction process for determining surgical methods can be implemented once the training process is complete.

Abstract: An information processing program causes a computer to execute a process including: generating, by inputting a captured image into a machine learning model, skeleton information on a person included in the captured image; detecting, by using the skeleton information, a specific motion of the person related to an object included in the captured image; specifying, by using positional information on the person included in the captured image, a first area in which the person is located at a time of detection of the specific motion from among a plurality of areas; specifying, by reading setting information recorded in a memory, first setting information that is associated with the first area; and identifying, based on the first setting information, a priority level of a notification of an alert related to the specific motion of the person related to the object.

Abstract: An apparatus and method for training and using a computing operation for digital image processing are provided. The apparatus and method may be used for 3-dimensional medical images. An exemplary method for digital image processing comprises: receiving an image displaying at least one detectable structure, determining the detectable structure; segmenting the image to obtain a segmentation mask that is associated with a geometric shape and comprises at least one quantifiable visual feature; generating a mesh based on the quantifiable visual feature; computing at least on quantifiable visual parameter based on the mesh; extracting quantifiable visual data from the image based on the quantifiable visual parameter; training the computing operation with the quantifiable visual data.

Abstract: A method for generating virtual microwave (MW) phantoms at a specific MW frequency or multiple MW frequencies, comprising generating virtual MW phantoms using a generative neural network (generator) and identifying the authenticity of the virtual MW phantoms using a discriminative neural network (discriminator).

Abstract: A detection and communication module arranged to implement a face detection function and an optical wireless communication function, and including a processing unit, a transmission chain and a reception chain, the processing unit being arranged to transmit via the transmission chain a detection signal, to receive via the reception chain the detection signal following its reflection on a face surface of an individual, and to evaluate a distance between the detection and communication module and the face surface of the individual, the processing unit being further arranged to transmit via the transmission chain a transmitted optical wireless communication signal containing data to be transmitted, and to receive via the reception chain a received optical wireless communication signal.

Abstract: Disclosed herein is a system and method for improving the accuracy of an object detector when trained with a dataset having a significant number of missing annotations. The method uses a novel Background Recalibration Loss (BRL) which adjusts the gradient direction according to its own activation to reduce the adverse effect of error signals by replacing the negative branch of the focal loss with a mirror of the positive branch when the activation is below a confusion threshold.