Abstract: Described is a system for object recognition. The system generates a training image set of object images from multiple image classes. Using a training image set and annotated semantic attributes, a model is trained that maps visual features from known images to the annotated semantic attributes using joint sparse representations with respect to dictionaries of visual features and semantic attributes. The trained model is used for mapping visual features of an unseen input image to its semantic attributes. The unseen input image is classified as belonging to an image class, and a device is controlled based on the classification of the unseen input image.

Abstract: A method for dividing, by a training system, a computational training work load of one or more neural network layers; pre-training the one or more neural network layers with a first class of image data sensitive to an original known dataset; generating a first weight file from the first layer of the neural network based on the first class of image data sensitive to the original known dataset; loading the one or more pre-trained neural network layers and the generated first weight file into at least one Internet of Things (IoT) device; stacking the one or more pre-trained neural network layers with the first layer of the neural network to form a new training system for an uploaded new dataset; adjusting the generated first weight file based on an input of one or more new classes of image data comprised in the uploaded new dataset to generate a new second weight file; inferencing an object class of new image data comprised on the uploaded new dataset using the generated new second weight file; and outputting th

Abstract: A method comprises (a) collecting (i) a set of chest computed tomography angiography (CTA) images scanned in the axial view and (ii) a manual segmentation of the images, for each one of multiple organs; (b) preprocessing the images such that they share the same field of view (FOV); (c) using both the images and their manual segmentation to train a supervised deep learning segmentation network, wherein loss is determined from a multi-dice score that is the summation of the dice scores for all the multiple organs, each dice score being computed as the similarity between the manual segmentation and the output of the network for one of the organs; (d) testing a given (input) pre-processed image on the trained network, thereby obtaining segmented output of the given image; and (e) smoothing the segmented output of the given image.

Abstract: Method for identifying a person through an electrocardiogram, ECG, waveform, said method comprising: capturing ECG signals from a sample population including the person to be identified; computing sample population ECG distances ST, RT and QT from the captured ECG signals; training a computer classification model on the computed sample population ECG distances, provided that no other ECG distances are used; capturing an ECG signal from the person to be identified; computing the person's ECG distances ST, RT and QT from the person's captured ECG signal; using the classification model with the person's computed ECG distances to identify the person to be identified within the sample population. Device for identifying a person through an electrocardiogram, ECG, waveform, said device comprising means for carrying out said method.

Abstract: An example of an apparatus including an input device to receive details about an event is provided. The apparatus further includes a camera to capture image data at the event. The apparatus also includes a memory storage unit to store the image data captured by the camera. In addition, the apparatus includes a reminder generator to generate a plurality of reminders during the event based on the details. Each reminder of the plurality of reminders requests an image from the camera. Furthermore, the apparatus includes a processor to index the image data in a database on the memory storage unit. The database is to be used to generate a gallery about the event.

Abstract: An imaging system for a smartphone includes a housing configured to rigidly attach to a handheld smartphone device. The system further includes an arm extending from the housing, wherein the arm is configured to be adjustable such that the arm is movable toward and away from the housing. Additionally, the system includes a sample mounting portion attached to the arm, the mounting portion having a retaining device, wherein the retaining device is configured to secure a target sample, the mounting portion configured to direct an image of the target sample to a first camera in the handheld smartphone device.

Abstract: Systems, methods, and storage media for automatically identifying skin conditions based on images and other data using machine learning methods and algorithms that encode expert knowledge of skin conditions. The systems, methods, and storage media may also leverage infrastructure that facilitates the collection, storage, labeling, dynamic organization, and expert review of existing and incoming data.

Abstract: An image processing system for reducing a calculation amount when calculating a parallax using an image of stereo photography. Reduced image data is generated by reducing image data respectively photographed by a plurality of cameras for stereo photographing, followed by reducing the number of pixels of image data. The image data and the reduced image data are stored in a memory. A plurality of reduced image data having the same photographing period are compared with each other to find a parallax. A parallax calculation region is set which includes the same position as a region where a parallax has been detected on reduced image data among a plurality of image data having the same photographing period. A parallax is calculated in the parallax calculation region of the image data formed before image reduction, and such parallax is used as a parallax of the image data formed before image reduction.

Abstract: The embodiments of the disclosure set a condition when image data is generated, in association with a position of image data. An image processing device makes a plurality of light emission states of a plurality of light emitting units different according to an illumination condition, and acquires a plurality of pieces of image data obtained by imaging in the plurality of different light emission states. The image processing device generates image data to be used for image measurement from a plurality of pieces of acquired image data on the basis of a generation condition defined in association with a position in the image data. In the image processing device, at least one of the illumination condition and the generation condition is determined so that the generated image data becomes image data suitable for a purpose of the image measurement.

Abstract: A system and method for tagging an image of an individual in a plurality of photos is disclosed herein. A feature vector of an individual is used to analyze a set of photos on a social networking website such as Facebook® to determine if an image of the individual is present in a photo of the set of photos. Photos having an image of the individual are tagged preferably by listing a URL or URI for each of the photos in a database.

Abstract: A computing-device implemented system and method for identifying an item in an x-ray image is described. The method includes training a machine learning algorithm with at least one training data set of x-ray images to generate at least one machine-learned model. The method further includes receiving at least one rendered x-ray image that includes an item, identifying the item using the at least one model, and generating an automated detection indication associated with the item.

Abstract: Some embodiments are directed to a monitoring system for monitoring the positioning of a patient relative to a treatment apparatus operable to deliver radiation along a radiation path. A model generation module is arranged to generate a model of a surface of a patient being monitored. A comparison module determines a transformation required to match the generated model to the stored model target surface. An offset determination module then utilizes the transformation and data indicative of a current position of a treatment apparatus to determine an offset value indicative of the offset of the surface of a patient relative to an axis corresponding to a determined current radiation path for radiation generated by the treatment apparatus applied to a patient. If the offset value exceeds a threshold, a warning may be generated or treatment may be halted.

Abstract: In accordance with the teachings described herein, systems and methods are provided for generating images for use in systems, e.g., imaging systems. The method includes receiving at least a first set of images, providing a first registration, providing a display, and displaying a first image on said display. Further, the method includes providing a user interface, providing a second registration, and displaying a second image in said user interface. Further, the systems include an image database, a display, and a registration engine. The registration engine includes software instructions stored in at least one memory device and executable by one or more processors.

Abstract: Disclosed embodiments provide techniques for estimating crowd flow rate. Video frames are obtained from a crowd scene. Groups of consecutive video frames form a time window. Crowd flow rate through the field of view of the camera is analyzed by estimating a total number of people contained within at least one frame of a time window. A subset of the people within the frame are identified as a group of tracker candidates. The tracker candidates are monitored throughout the time window, and the time when the tracker candidates exit the field of view of the camera is used to estimate a crowd flow rate for the number of people passing through the field of view of the camera in a given time interval. Thus, disclosed embodiments can improve the technical field of crowd flow rate estimation.

Abstract: An image processing training method includes the following steps. A template label image is obtained, in which the template label image comprises a label corresponding to a target. A plurality of first reference images are obtained, in which each of the first reference images comprises object image data corresponding to the target. A target image according to the template label image and the first reference images is generated, in which the target image comprises a generated object, a contour of the generated object is generated according to the template label image, and a color or a texture of the target image is generated according to the first reference images.

Abstract: An adaptive method for generating CT image data is described. In the method, projection measurement data of an examination region of an examination object is acquired. Furthermore, uncorrected image data of the examination region is generated. Artifact-affected subregions of the examination region are determined on the basis of at least one part of the uncorrected image data. An artifact-reduced image reconstruction is carried out in the artifact-affected subregions of the examination region. Only artifact-reduced subimage data of the artifact-affected subregions is generated. Finally, artifact-reduced image data of the entire examination region is generated by combining at least one part of the uncorrected image data and the artifact-reduced subimage data. A reconstruction device is also described. Moreover, a computed tomography system is described.

Abstract: A driving consciousness estimation device is configured to estimate a driver's driving consciousness of a host vehicle as a driving readiness degree. The device includes a viewing target recognition unit configured to recognize a viewing target used for estimating the driving readiness degree based on a result of detection by an external sensor of the host vehicle, a driver's line of sight recognition unit configured to recognize a driver's line of sight, a viewing reaction index calculation unit configured to calculate a driver's viewing reaction index for the viewing target based on the driver's line of sight recognized by the driver's line of sight recognition unit, and a driving readiness degree estimation unit configured to estimate the driving readiness degree based on the viewing reaction index.

Abstract: An image processing method transforms image sequences into luminances, filters the luminances, determines the temporal differences between the luminances, performs a frequency domain transformation on the temporal differences, and applies a temporal contrast sensitivity function envelope integral to the frequency transform output to generate a temporal image metric. The temporal image metric may be applied for example to train a neural network or to configure a display device to depict a visual indication of the temporal image metric.

Abstract: Systems and methods for estimating an object of fixation from a gaze of a driver. The method includes: receiving image data from a plurality of input devices; processing the image data from a first one of the plurality of input devices to identify an object track; analyzing the image data from a second one of the plurality of input devices and the image data from the first one of the plurality of input devices to determine a projected gaze of a driver; analyzing the object track and the projected gaze to identify a plurality of objects in the gaze of the driver; performing a probability analysis to estimate the object of fixation from among the plurality of objects; and generating an output image identifying the estimated object of fixation.

Abstract: An information processing apparatus includes a generating unit configured to generate a reference image based on an image of a first object, an image of a second object, and processing information on processes including a process of inserting a second object into the first object. The first object has at least one window portion through which the inserted second object is at least partially visible, a verifying unit configured to verify a captured image with the generated reference image, the captured image being obtained by capturing the first object into which the second object has been inserted, and an outputting unit configured to output a result of the verification.