Abstract: A method for generating a summary video includes generating a user emotion graph of a user watching a first video. The method also includes obtaining a character emotion graph for a second video, by analyzing an emotion of a character in a second video that is a target of summarization. The method further includes obtaining an object emotion graph for an object in the second video, based on an object appearing in the second video. Additionally the method includes obtaining an image emotion graph for the second video, based on the character emotion graph and the object emotion graph. The method also includes selecting at least one first scene in the second video by comparing the user emotion graph with the image emotion graph. The method further includes generating the summary video of the second video, based on the at least one first scene.

Abstract: Techniques are described for determining cognitive impairment, an example of which includes accessing a set of epochs of magnetoencephalography (MEG) data of responses of a brain of a test patient to a plurality of auditory stimulus events; processing the set of epochs to identify parameter values one or more of which is based on information from the individual epochs without averaging or otherwise collapsing the epoch data. The parameter values are input into a model that is trained based on the parameters to determine whether the test patient is cognitively impaired. Graphical user interfaces are described for presenting MEG epoch data and a score that correlates to a likelihood of the test individual being cognitively impaired.

Abstract: A method and system of two-point correction based on temperature substitution, includes: a black body related to the lens temperature is arranged outside a lens, a first black body arranged in front of the lens of an infrared thermal camera, so as to collect a gray value of the first black body; a temperature of the first black body being equal to the lens temperature; a second black body arranged in front of the lens of the infrared thermal camera, so as to collect a gray value of the second black body; a temperature of the second black body being higher than the lens temperature. Two correction parameters obtained by black body calculation related to temperature are strongly applicable, which eliminates the problem of image quality deterioration caused by temperature changes and greatly eliminates the pot lid phenomenon due to non-uniformity, thereby accurately calculating the corrected gray value.

Abstract: In a case where it is identified based on fisheye information that an input image is a fisheye image, a block position computing section computes the position of a processing target block relative to a fisheye center, and outputs, to a table selecting section, positional information representing the computed position. In a case where it is identified based on fisheye information that an input image is a fisheye image, the table selecting section selects, on the basis of positional information, an intra-prediction mode table according to the position of a processing target block from a plurality of intra-prediction mode tables, and outputs the selected intra-prediction mode table to a prediction image generating section.

Abstract: An image processing device includes an estimator (54) configured to estimate a degree of fogging of image data on the basis of the image data, the image data having been acquired by an imaging device (41) capturing an image of an imaging target (T) disposed in a casing (25) of an air treatment device (10); and a determiner (55) configured to determine output image data on the basis of the degree of fogging of the image data estimated by the estimator (54), the output image data being image data to be output.

Abstract: The present disclosure describes techniques of detecting objects in a video. The techniques comprises extracting features from each frame of the video; generating a first attentive feature by applying a first attention model on at least some of features extracted from any particular frame among the plurality of frames, wherein the first attention model identifies correlations between a plurality of locations in the particular frame by computing relationships between any two locations among the plurality of locations; generating a second attentive feature by applying a second attention model on at least one pair of features at different levels selected from the features extracted from the particular frame, wherein the second attention model identifies a correlation between at least one pair of locations corresponding to the at least one pair of features; and generating a representation of an object included in the particular frame.

Abstract: Methods and systems to improve a visual perception of dark scenes in video. An example device includes one or more processors to receive a frame of video segmented into a plurality of sub-regions. A local luminance histogram is generated for each sub-region. A global luminance histogram is generated for the entire frame of video and a global tone mapping curve is generated based on the global luminance histogram. A tone mapping LUT is generated for each sub-region based on the global tone mapping curve and the corresponding local luminance histogram for the sub-region. The frame of video is then modified using the tone mapping LUTs generated for each sub-region and sent to an output device.

Abstract: Systems and methods for designing and implementing patient-specific surgical procedures and/or medical devices are disclosed. An example method includes obtaining, first multi-dimensional image data of an anatomical region of a patient in a first loading state; determining, from the first multi-dimensional image data, regions corresponding to anatomic elements of a spine; identifying, in each region, a first set landmarks; obtaining, a second multi-dimensional image data of the anatomical region comprising the spine of the patient in a second loading state; identifying a second set of landmarks from the second multi-dimensional image data that map to the first set of landmarks; obtaining an aligned multi-dimensional image data of the patient's spine; generating a design for a medical implant based on spinopelvic parameters measured from the aligned multi-dimensional image data; and causing a medical implant to be manufactured by sending the design for the medical implement to a manufacturing device.

Abstract: Examples include detecting objects and determining a set of features for the objects. Examples include receiving a first image input, generating a number of feature maps from the first image input using a number of convolution filters, generating a first number of fully connected layers directly based on the number of feature maps, and detecting a number of objects in the first image and determining a set of features for each object from the first number of fully connected layers.

Abstract: A physiological state determination device determines a predetermined physiological state of a subject. The physiological state determination device includes a brain function activation information detection unit, a face change information acquisition unit, and a physiological state determination unit. The brain function activation information detection unit detects brain function activation information corresponding to a physiological state. The face change information acquisition unit acquires face change information indicating a time-series change in face data of a subject. The physiological state determination unit determines the predetermined physiological state of the subject based on the brain function activation information and the face change information.

Abstract: Optical center is determined on a column-by-column and row-by-row basis by identifying brightest pixels in respective columns and rows. The brightest pixels in each column are identified and a line is fit to those pixels. Similarly, brightest pixels in each row are identified and a second line is fit to those pixels. The intersection of the two lines is the optical center.

Abstract: A method and system for generating material decomposition images from plural-energy x-ray based imaging, the method comprising: modelling spatial relationships and spectral relationships among the plurality of images by learning features from the plurality of images in combination and one or more of the plurality of images individually with a deep learning neural network; generating one or more basis material images employing the spatial relationships and the spectral relationships; and generating one or more material specific or material decomposition images from the basis material images. The neural network has an encoder-decoder structure and includes a plurality of encoder branches; each of one or more of the plurality of encoder branches encodes two or more images of the plurality of images in combination; and each of one or more of the plurality of encoder branches encodes a respective individual image of the plurality of images.

Abstract: The present invention discloses a large-field-angle image real-time stitching method based on calibration. First, a calibration algorithm is used to solve the positional relationship between cameras, and the prior information is used to solve a homography matrix between images. The system is easy to build, and the program is simple and easy to implement; an overlapping area ROI of images can be calculated by the homography matrix between images, and an energy model thereof can be built and solved with a graph cut algorithm; the graph cut algorithm has high time complexity and depends on the number of nodes in a graph; here, images are divided into layers, and solutions are obtained layer by layer and iterated; and finally, a stitched image is further optimized by simple linear fusion of stitching seams and histogram equalization of the stitched image.

Abstract: Systems and methods of generating an orthodontic model are disclosed. The method may include: generating an initial model of a patient dentition; generating a target model of the patient dentition; defining a plurality of caps and a plurality of links, wherein each link connects two of the plurality of caps; generating a relaxed model of a dental appliance from the plurality of caps and the plurality of links; generating a deformed model of a dental appliance from the plurality of caps and plurality of links; and determining a plurality of movements, wherein the plurality of moments transform the relaxed model to the deformed model and wherein the moments are configured to direct the patient dentition from the initial model to the target model.

Abstract: Techniques are described for generating reformatted views of a three-dimensional (3D) anatomy scan using deep-learning estimated scan prescription masks. According to an embodiment, a system is provided that comprises a memory that stores computer executable components, and a processor that executes the computer executable components stored in the memory. The computer executable components comprise a mask generation component that employs a pre-trained neural network model to generate masks for different anatomical landmarks depicted in one or more calibration images captured of an anatomical region of a patient. The computer executable components further comprise a reformatting component that reformats 3D image data captured of the anatomical region of the patient using the masks to generate different representations of the 3D image data that correspond to the different anatomical landmarks.

Abstract: Techniques for improved camera calibration are disclosed. An image is analyzed to identify a first set of key points for an object. A virtual object is generated. The virtual object has a second set of key points. A reprojected version of the second set is fitted to the first set in 2D space until a fitting threshold is satisfied. To do so, a 3D alignment of the second set is generated in an attempt to fit (e.g., in 2D space) the second set to the first set. Another operation includes reprojecting the second set into 2D space. In response to comparing the reprojected second set to the first set, another operation includes determining whether a fitting error between those sets satisfies the fitting threshold. A specific 3D alignment of the second set is selected. The camera is calibrated based on resulting reprojection parameters.

Abstract: A volumetric segmentation method is disclosed for brain region analysis, in particular but not limited to, regions of the basal ganglia such as the subthalamic nucleus (STN). This serves for visualization and localization within the sub-cortical region of the basal ganglia, as an example of prediction of a region of interest for deep brain stimulation procedures. A statistical shape model is applied for variation modes of the STN, or the corresponding regions of interest, and its predictors on high-quality training sets obtained from high-field, e.g., 7T, MR imaging. The partial least squares regression (PLSR) method is applied to induce the spatial relationship between the region to be predicted, e.g., STN, and its predictors. The prediction accuracy for validating the invention is evaluated by measuring the shape similarity and the errors in position, size, and orientation between manually segmented STN and its predicted one.

Abstract: For an easy calibration using calibration particles, provided is a measuring device to capture images of target objects. An image analyzer acquires multiple images obtained at a predetermined time interval, (a) specifies the mean-square displacement of a bright point of a calibration particle based on the displacement of the bright point of the calibration particle in the multiple images in a calibration mode, and (b) specifies the mean-square displacement of a bright point of the target particle based on the displacement of the bright point of the target particle in the multiple images in a measurement mode. A particle size analyzer (c) derives the particle size of the target particle from the mean-square displacement of the bright point of the target particle based on the mean-square displacement of the bright point of the calibration particle and the particle size of the calibration particle in an analysis mode.

Abstract: A method of quantitatively evaluating a cognitive status and its future change from a medical image of an individual's brain, the method comprising scanning the individual's brain with a scanning device so as to acquire at least one medical brain image; processing the medical brain image to obtain at least one feature of the image; using a pre-established prediction model to determine a condition of the cognitive status and predict its future change based on the at least one feature obtained.

Abstract: A method and apparatus for removing breathing motion artifacts in imaging CT scans is disclosed. The method acquires raw imaging data from a CT scanner, and processes the raw CT imaging data by removing motion-induced artifacts via a motion model. Processing the imaging data may be achieved by initially estimating a 3D image to provide an estimate of raw sinogram image data, comparing the estimate to an actual CT sinogram, determining a difference between the sinograms, and iteratively reconstructing the 3D image by using the difference to alter the 3D image until the sinograms agree, wherein the 3D image moves according to the motion model.