Abstract: An approach is provided for determining quality assessment of cross view feature correspondence using bundle adjustment techniques. The approach, for example, involves retrieving a plurality of annotated images that are labeled with one or more feature correspondence labels. The approach also involves performing a bundle adjustment process on the plurality of annotated images to compute a three-dimensional (3D) location and a residual error of a feature corresponding to the one or more feature correspondence labels. The approach further involves flagging the one or more feature correspondence labels as potentially incorrect based on determining that the residual error is greater than an error threshold.

Abstract: The present invention relates to a method of assisting in diagnosis of a target heart disease using a retinal image, the method including: obtaining a target retinal image which is obtained by imaging a retina of a testee; on the basis of the target retinal image, obtaining heart disease diagnosis assistance information of the testee according to the target retinal image, via a heart disease diagnosis assistance neural network model which obtains diagnosis assistance information that is used for diagnosis of the target heart disease according to the retinal image; and outputting the heart disease diagnosis assistance information of the testee.

Abstract: A computer-implemented method of performing a convolution between an input data array and a kernel to generate an output data array includes decomposing the kernel into a plurality of sub-kernels each having a respective position relative to the kernel and respective in-plane dimensions less than or equal to a target kernel dimension, and for each of the plurality sub-kernels: determining a respective portion of the input data array on the basis of the respective in-plane dimensions of the sub-kernel and the respective position of the sub-kernel relative to the kernel; retrieving the respective portion of the input data array; and performing a convolution between the retrieved respective portion of the input data array and the sub-kernel to generate a respective intermediate data array. The method further includes summing the generated intermediate data arrays to generate at least a portion of the output data array.

Abstract: A system and method for generating a rib suppressed radiographic image using deep learning computation. The method includes using a convolutional neural network module trained with pairs of a chest x-ray image and its counterpart bone suppressed image. The bone suppressed image is obtained using a bone suppression algorithm applied to the chest x-ray image. The convolutional neural network module is then applied to a chest x-ray image or the bone suppressed image to generate an enhanced bone suppressed image.

Abstract: Provided herein are orthodontic devices and methods for patients with missing or ectopic teeth. Methods and processes are provided to properly number the teeth of a patient's arch after a dental scan. Methods and processes are also provided to automatically detect missing or ectopic teeth after a dental scan. Methods of designing and manufacturing the aligner are also provided.

Abstract: Systems and methods are described, and an example system creates a transaction between a biometric station and a user in a group of users. Transaction includes establishing for the user, a ground-truth identity, capturing an image of the user in the biometric station, making a first determination of whether the image is sufficient for identification of the user. When the image is sufficient, the transaction derives biometric information from the image, converts the biometric information into a generic-form biometric information, agnostic of how the biometric information was obtained, and generated hashed user data by applying a hash function to the generic-form biometric information. The transaction makes a second determination, based on the hashed user data, of whether the user is represented in a gallery of sample images. The transactions are repeated for other users in the group.

Abstract: A method may include obtaining scan data that is captured by scanning a heart of a subject with the imaging device. The method may include obtaining electrocardiogram (ECG) data during the scans of the heart of the subject. The method may include determining first scan data from the scan data. The method may include reconstructing the cardiac image based on the first scan data. The method may also include performing one or more iterations. Each of the one or more iterations may include determining whether there is a pulsatile artifact in the reconstructed cardiac image, determining a second time period of at least one cardiac cycle based on the pulsatile artifact, and reconstructing the cardiac image based on the scan data and the second time period.

Abstract: Embodiments of the present disclosure provide a method and an apparatus for generating a simulation scene. The method includes: acquiring scene parameters of a benchmark scene, where a dimensionality of the scene parameters of the benchmark scene is M; inputting the scene parameters of the benchmark scene into an encoder that is trained, and acquiring encoding parameters according to an output result of the encoder, where a dimensionality of the encoding parameters is N, and N<M; adjusting the encoding parameters to obtain adjusted encoding parameters, and inputting respectively the adjusted encoding parameters into a decoder that is trained; and generating a simulation scene according to the scene parameters of the reconstructed scene that are output by the decoder. According to the method, generating massive simulation scenes similar to the benchmark scene based on the benchmark scene can be achieved, which meets the diversity requirements for the simulation scenes.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training an event detector. The methods, systems, and apparatus include actions of obtaining frames of a video, determining whether an object of interest is detected within the frames, determining whether motion is detected within the frames, determining whether the frames correspond to motion by an object of interest, generating a training set that includes labeled inter-frame differences based on whether the frames correspond to motion by an object of interest, and training an event detector using the training set.

Abstract: An exemplary process for identifying a type of a physical environment amongst a plurality of types of physical environments is provided. The process includes obtaining, using the one or more cameras, image data corresponding to a physical environment. The process further includes identifying at least one portion of an entity in the physical environment based on the image data; determining, based on the identified at least one portion of the entity, whether the entity is an entity of a first type; determining a type of the physical environment if the entity is an entity of the first type; and presenting one or more virtual objects and a representation of the entity.

Abstract: An image processing apparatus that derives an optical flow which is a set of motion vectors of at least a part of pixels between a plurality of input images includes an acquisition unit and a deriving unit. The acquisition unit is configured to acquire a contrast value of one of the input images. The deriving unit is configured to derive the optical flow based on the contrast value acquired by the acquisition unit and an energy function. The energy function includes a data term specifying a correlation between the plurality of input images and a smoothing term specifying a degree of smoothing.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for classifying actions of an object. The methods, systems, and apparatus include actions of: obtaining frames of video including an object of interest; determining a type of action of the object in each of the frames of video; determining a group of frames from the frames of video based on the type of action; determining an aggregated background subtraction (ABS) image based on adjacent frames of the group of frames; generating a training set that includes labeled ABS images including the ABS image; and training an action classifier using the training set.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for obtaining a foreground occupancy map for a camera view. The methods, systems, and apparatus include actions of determining an area of an image in which there is a false detection of an object, determining a likely contribution of the area to the false detection based on the foreground occupancy map, generating a modified object detector based on the likely contribution of the area, and detecting an object using the modified object detector.

Abstract: A superimposing position correction device includes an image acquisition unit, a difference image generation unit, a processing plan image generation unit, a subregion generation unit, a similar shape search unit, a representative point extraction unit, a projection matrix calculation unit and a superimposition display unit.

Abstract: A striped pattern image examination support apparatus includes a feature extraction part, a central line collation part, and a display part. The feature extraction part extracts, from each of a first striped pattern image and a second striped pattern image, at least central lines and feature points, as a feature of each of the first striped pattern image and the second striped pattern image. The central line collation part performs collation of the respective central lines of the first striped pattern image and the second striped pattern image, and computes corresponding central lines between the first striped pattern image and the second striped pattern image. The display part determines a display form of each of the central lines based on the computed corresponding central lines and superimposes and displays the central lines on each of the first striped pattern image and the second striped pattern image, according to the determined display form.

Abstract: The disclosure herein describes training a machine learning model to recognize a real-world object based on generated virtual scene variations associated with a model of the real-world object. A digitized three-dimensional (3D) model representing the real-world object is obtained and a virtual scene is built around the 3D model. A plurality of virtual scene variations is generated by varying one or more characteristics. Each virtual scene variation is generated to include a label identifying the 3D model in the virtual scene variation. A machine learning model may be trained based on the plurality of virtual scene variations. The use of generated digital assets to train the machine learning model greatly decreases the time and cost requirements of creating training assets and provides training quality benefits based on the quantity and quality of variations that may be generated, as well as the completeness of information included in each generated digital asset.

Abstract: A method and an apparatus for generating a High Dynamic Range, HDR, image are proposed. The method comprises obtaining a set of two or more input images, the two or more input images including a reference image and one or more non-reference images; for each of the one or more non-reference images, performing an image analysis which comprises, for each region of a plurality of regions of the non-reference image, assessing whether the region of the non-reference image and a corresponding region of the reference image show the same image content and declaring the region of the non-reference image as valid or as invalid based on the assessment; and generating the HDR image by fusing the reference image and the one or more non-reference images, wherein the fusing comprises, for each of the one or more non-reference images, disregarding the invalid regions of the respective non-reference image.

Abstract: A method for automatically processing at least one image slice of a given part of at least one lung of a patient suffering from a pathology that causes a bronchial infection by diffuse dilatation of the bronchial tubes of the lungs.

Abstract: This disclosure relates to a method and system for summarizing multimedia content. The method may include receiving multimedia content. The method may further include identifying one or more primary objects, wherein identifying the primary objects comprises identifying one or more actions associated with the primary objects and one or more interactions between the primary objects and one or more secondary objects. The primary objects are associated with one or more parameters. The method may further include determining at least one primary object of interest from the primary objects by selectively prioritizing the parameters. The method may further include summarizing the multimedia content based on the primary object of interest, actions associated with the primary object of interest, interactions between the primary object of interest and the secondary objects, and interactions between the secondary objects and one or more tertiary objects.

Abstract: In an approach for classifying an image containing a continuous characteristic using an image classifier, an image tagged with a single value is received, wherein the single value is from a set of ordered value. A processor sets at least two thresholds based on the set of ordered values. A processor derives at least two labels according to the single value tagged in the image and the at least two thresholds. A processor runs image through an image classifier. A processor receives at least two predictions from the image classifier based on the at least two labels. A processor adjusts the image classifier based on differences between the at least two predictions and the at least two labels.