Abstract: A system for on-vehicle camera alignment monitoring includes an on-vehicle camera in communication with a controller. The controller monitors vehicle operating parameters and camera signal parameters, and captures an image file from the on-vehicle camera. A first level analysis of the image file, the vehicle operating parameters, and the camera signal parameters is executed to detect dynamic conditions and image feature parameters that affect camera alignment. An error with one of the dynamic conditions or the image feature parameters that affects the camera alignment is detected. A second level analysis of the camera signal parameters is executed to identify a root cause indicating one of the dynamic conditions or the image feature parameters that affects the camera alignment based upon the error. A camera alignment-related fault is detected based upon the root cause, and vehicle operation is controlled based upon the camera alignment-related fault.

Abstract: We provide an imaging system for use in a fluoroscopy procedure carried out on a subject, the imaging system being configured to receive images of a portion of the subject from an image acquisition device, the imaging system comprising an interface module for displaying images received from the image acquisition device to a user, and an optimisation module for determining, based on an image received from the image acquisition device, one or more properties of the image, wherein the optimisation module is configured to output a control routine.

Abstract: Systems, methods, and computer program products for determining an attack on a neural network. A data sample is received at a first classifier neural network and at a watermark classifier neural network, wherein the first classifier neural network is trained using a first dataset and a watermark dataset. The first classifier neural network determines a classification label for the data sample. A watermark classifier neural network determines a watermark classification label for the data sample. A data sample is determined as an adversarial data sample based on the classification label for the data sample and the watermark classification label for the data sample.

Abstract: Disclosed herein include systems and methods for biological object tracking and lineage construction. Also disclosed herein include cloud-based systems and methods for allocating computational resources for deep learning-enabled image analysis of biological objects. Also disclosed herein include systems and methods for annotating and curating biological object tracking-specific training datasets.

Abstract: An encoded substrate to be filmed by a camera device for generating an image is provided. The encoded substrate includes a plurality of grids arranged in a form of two-dimensional array, wherein each grid includes a first pattern and a second pattern not overlapped with each other. The first pattern corresponds to a first-dimensional encoded value, and the second pattern corresponds to a second-dimensional encoded value. The image is processed by a processor for scanning the plurality of grids. In a first-dimensional direction, the processor outputs a first coordinate according to at least two first patterns corresponding to at least two consecutive grids among the plurality of grids. In a second-dimensional direction, the processor outputs a second coordinate according to at least two second patterns corresponding to at least two consecutive grids among the plurality of grids.

Abstract: An information processing device and an information processing method capable of reducing an information processing amount of an application processor that recognizes a subject from an input image are provided. An information processing device according to the present disclosure includes an acquisition unit, a generation unit, and an output unit. From an imaging unit that captures an image and generates image data, the acquisition unit acquires the image data. The generation unit generates, from the image data acquired by the acquisition unit, metadata to assist an application processor that executes processing related to the image. The output unit outputs the metadata generated by the generation unit to the application processor.

Abstract: A tool for providing contextual policy-based computer vision control. The tool identifies one or more camera devices in an environment. The tool determines a plurality of vision areas in a field of view within the environment. The tool determines one or more contextual policies for the plurality of vision areas in the field of view. The tool determines one or more vision augmentations for the plurality of vision areas based, at least in part, on an aggregate computer vision capability of the one or more camera devices in the environment and the one or more contextual policies for the plurality of vision areas in the field of view. The tool applies the one or more vision augmentations to at least one of the one or more camera devices in the environment.

Abstract: In an object tracking device, an object detection unit detects an object from video data using a plurality of object detection programs of different detection accuracy. The object tracking unit tracks the object from the detected position of the object by the object detection unit. The control unit controls the object detection programs used by the object detection unit based on the detection result of the object by the object detection unit and the tracking result of the object by the object tracking unit.

Abstract: The disclosure provides a reversible watermarking method for the oblique photography three-dimensional models with controllable accuracy. According to the disclosure: firstly, by using the global stability of the average included angle of vertex normal vector, the feature points of the oblique photography three-dimensional model are extracted; secondly, the mapping relationship is established by the ratio of the distance between feature points and non-feature points, and the vertexes are grouped, and each group consists of a feature point corresponding to several non-feature points; finally, the spherical coordinate system is constructed by taking the feature points as the coordinate origin in grouping, and the watermark is embedded by modifying the radius of the coordinate system.

Abstract: Techniques are described for tailoring automatic exposure control (AEC) settings to specific patient anatomies and clinical tasks. According to an embodiment, computer-implemented method comprises receiving one or more scout images captured of an anatomical region of a patient in association with performance of a computed tomography (CT) scan. The method further comprises employing a first machine learning model to estimate, based on the one or more scout images, expected organ doses representative of expected radiation doses exposed to organs in the anatomical region under different AEC patterns for the CT scan. The method can further comprises employing a second machine learning model to estimate, based on the one or more scout images, expected measures of image quality in target and background regions of scan images captured under the different AEC patterns, and determining an optimal AEC pattern based on the expected organ doses and the expected measures of image quality.

Abstract: A device may obtain first information relating to one or more first lung nodules identified in first imaging of a chest of a patient and second information relating to one or more second lung nodules identified in second imaging of the chest of the patient. The device may provide the first information and the second information to a machine learning model. The device may determine, using the machine learning model, a risk of lung cancer associated with the patient based on an elapsed time between performance of the first imaging and the second imaging and differences between the first information and the second information. The risk of lung cancer may have an inverse correlation to the elapsed time and a direct correlation to the differences.

Abstract: A method for processing an image uses a generator which is a machine learning model. The generator converts an input low resolution image into a first feature map. From the first feature map, a first intermediate image and a second intermediate image each having resolution higher than resolution of the low resolution image are generated. Based on the first intermediate image and the second intermediate image, an estimated image having higher resolution than the resolution of the low resolution image is generated.

Abstract: An apparatus and a method correct a level of a headlamp, in which a sight of the driver may be sufficiently secured while an influence on a sight of a driver of a preceding vehicle is minimized by specifying a level of a cutoff line of the headlamp using a transverse length and a longitudinal length of a number plate frame image. The specified level is compared with a level of a target line calculated based on a change of an inter-vehicle distance. The method includes determining whether it is necessary to perform a correction of raising or lowering the level of the headlamp, and a correcting the level of the headlamp such that the cutoff line of the headlamp follows the level of the target line at the corresponding inter-vehicle distance.

Abstract: The present invention is directed to a system and methods of predicting undiagnosed mental disorders in a patient and determining the best medication-class treatment for the mental disorder through the analysis of functional magnetic resonance imaging of functional connectivity of the brain representative of a certain mental disorder.

Abstract: A system including first and second camera assemblies, and a processor. The first camera assembly includes a first camera producing, in a first imaging modality, a first image of an organ acquired from a first angle, and a first position sensor, producing a first position signal indicative of a first position and a first orientation, of the first camera. The second camera assembly includes a second camera, producing, in a second imaging modality, a second image of the organ acquired from a second angle, and a second position sensor, producing a second position signal indicative of a second position and a second orientation, of the second camera. The processor registers between the first and second images based on the first and second position signals and displays, based on the first and second images, a third image including a combination of at least part of the first and second images.

Abstract: A computer device obtains a to-be-annotated image having a first magnification. The device obtains an annotated image from an annotated image set, the annotated image distinct from the to-be-annotated image and having a second magnification that is distinct from with the first magnification. The annotated image set includes at least one annotated image. The device matches the to-be-annotated image with the annotated image to obtain an affine transformation matrix, and generates annotation information of the to-be-annotated image according to the affine transformation matrix and the annotated image. In this way, annotations corresponding to images at different magnifications may be migrated. For example, the annotations may be migrated from the low-magnification images to the high-magnification images, thereby reducing the manual annotation amount and avoiding repeated annotations, and further improving annotation efficiency and reducing labor costs.

Abstract: Embodiments include a system for determining cardiovascular information for a patient. The system may include at least one computer system configured to receive patient-specific data regarding a geometry of the patient's heart, and create a three-dimensional model representing at least a portion of the patient's heart based on the patient-specific data. The at least one computer system may be further configured to create a physics-based model relating to a blood flow characteristic of the patient's heart and determine a fractional flow reserve within the patient's heart based on the three-dimensional model and the physics-based model.

Abstract: Various embodiments of the teachings herein include a building object detection system for a building comprising: a scanner to generate scan data; an artificial intelligence module implemented on a processing unit trained to process the scan data to detect and automatically classify building objects installed within the space; a database to store different types of building objects and associated installation rules; and a verification unit to verify the classification of a building object based on compliance with the installation rules. The AI is trained using labeled training data comprising labels derived from scan data of previously scanned building objects provided with associated object identifiers. A second set of installation rules is loaded from the database based at least in part on a location of the building.

Abstract: Embodiments include a system for determining cardiovascular information for a patient. The system may include at least one computer system configured to receive patient-specific data regarding a geometry of the patient's heart, and create a three-dimensional model representing at least a portion of the patient's heart based on the patient-specific data. The at least one computer system may be further configured to create a physics-based model relating to a blood flow characteristic of the patient's heart and determine a fractional flow reserve within the patient's heart based on the three-dimensional model and the physics-based model.

Abstract: A lane link generation device includes an acquisition unit, a lane link generation unit, and a connection unit. The connection unit acquires first and second sections representing vehicle traveling areas. The lane link generation unit generates lane links in the first section and the second section. When the lane link of the first section and the lane link of the second section have endpoints, the connection unit determines a connection destination endpoint which is an endpoint of a connection destination to be connected to an endpoint and which is an endpoint belonging to a different section, based on determination rule information. The connection unit generates a lane link that connects the endpoint to the determined connection destination endpoint.