Abstract: An electronic device includes a plurality of cameras, and at least one processor connected to the plurality of cameras. The at least one processor is configured to, based on a first user command to obtain a live view image, segment an image frame obtained via a camera among the plurality of cameras into a plurality of regions based on a brightness of pixels and an object included in the image frame; obtain a plurality of camera parameter setting value sets, each including a plurality of parameter values with respect to the plurality of regions; based on a second user command to capture the live view image, obtain a plurality of image frames using the plurality of camera parameter setting value sets and at least one camera among the plurality of cameras; and obtain an image frame by merging the plurality of obtained image frames.

Abstract: A dual variational autoencoder-generative adversarial network (VAE-GAN) is trained to transform a real video sequence and a simulated video sequence by inputting the real video data into a real video decoder and a real video encoder and inputting the simulated video data into a synthetic video encoder and a synthetic video decoder. Real loss functions and simulated loss functions are determined based on output from a real video discriminator and a simulated video discriminator, respectively. The real loss functions are backpropagated through the real video encoder and the real video decoder to train the real video encoder and the real video decoder based on the real loss functions. The synthetic loss functions are backpropagated through the synthetic video encoder and the synthetic video decoder to train the synthetic video encoder and the synthetic video decoder based on the synthetic loss functions.

Abstract: An automated line clearance inspection system will enable fast and accurate inspection of packaging equipment lines to reduce or prevent product intermixing. The system includes a set of image capturing devices that are controlled via a central processing unit whereby end run images are compared with control images to determine if a line is cleared.

Abstract: An information processing apparatus (10) according to the present disclosure includes: an object recognition unit (11) that outputs, by using a first modal signal and a first modal recognition model corresponding to the first modal signal, an inference result regarding the first modal signal; a training data processing unit (12) that generates first modal training data regarding the first modal signal by using the inference result, and updates second modal training data regarding a second modal signal that is different from the first modal signal by using the first modal training data; and a recognition model update unit (13) that updates a second modal recognition model corresponding to the second modal signal by using the second modal training data.

Abstract: Embodiments are disclosed for finding similar persons in images. In particular, in one or more embodiments, the disclosed systems and methods comprise receiving an image query, the image query including an input image that includes a representation of a person, generating a first cropped image including a representation of the person's face and a second cropped image including a representation of the person's body, generating an image embedding for the input image by combining a face embedding corresponding to the first cropped image and a body embedding corresponding to the second cropped image, and querying an image repository in embedding space by comparing the image embedding to a plurality of image embeddings associated with a plurality of images in the image repository to obtain one or more images based on similarity to the input image in the embedding space.

Abstract: This document describes techniques, apparatuses, and systems for enabling multispectral object-detection with thermal imaging. A thermal sensor, a multispectral stereo camera system, and an illumination system are used in combination to provide a multispectral object-detection system for use with safety or driving systems for autonomous and semi-autonomous vehicles. The illumination system can provide illumination for the stereo camera system and serve as a range-finder that can determine a distance between the stereo camera system and an object in the system's field-of-view. The stereo camera system can include one or more of various camera or sensor technologies, including infrared or near-infrared, thermal imaging, or visible light. The system can be used to detect objects in the field-of-view, determine a distance to the object, and estimate the object's size and relative motion.

Abstract: According to one embodiment, a biometric authentication device includes a sensor and an illumination device. The sensor includes a first base material, a plurality of photoelectric conversion elements and a first control unit. The illumination device includes a second base material, a plurality of light emitting elements and a second control unit. Upon lighting the light emitting element at a predetermined position in the groups of the light emitting elements by the second control unit, the first control unit controls the operation of the photoelectric conversion elements so as to acquire a signal from the photoelectric conversion elements at predetermined positions of the groups of the photoelectric conversion elements.

Abstract: Methods and systems are provided for establishing an activity zone within video captured by a camera. A user device may establish an activity zone or an exclusion zone within video representing a physical view captured by the camera. An automatic monitoring system may analyze the video of the camera with respect to the activity zone or exclusion zone and trigger notifications that are associated with a zone so as to limit notifications transmitted to the user device to those notifications that are relevant to the user. With activity or exclusion zones video from a camera may be automatically monitored to reduce a volume of notifications transmitted to the user device.

Abstract: An object detection method, for detecting a target object, comprising: capturing at least two detection portions with a first aspect ratio from an input image with a second aspect ratio; confirming whether any object is detected in each of the detection portions and obtaining corresponding boundary boxes for detected objects; and wherein the first aspect ratio is different to the second aspect ratio.

Abstract: A method for training an automatic learning machine (3.12), comprising the following steps: defining, in a database, at least one first class of image matrices (3.11) and one second class of image matrices (3.11), the first class of image matrices being associated with said first measured characteristic value; the second class of image matrices (3.11) being associated with said second measured characteristic value; classifying said image matrices in said classes as a function of the respective measured characteristic value for the mixture of plant elements from which each of said images has been obtained; modifying, by said calculator (3.15), in response to the reception of said overall classification error (3.18), at least one of its calculation elements, so that the second probabilities of inclusion of the image matrices in said at least two classes are closer to the true inclusion of the at least three image matrices in the at least two classes than the first inclusion probabilities.

Abstract: Various implementations disclosed herein include devices, systems, and methods that detect surfaces and reflections in such surfaces. Some implementations involve providing a CGR environment that includes virtual content that replaces the appearance of a user or the user's device in a mirror or other surface providing a reflection. For example, a CGR environment may be modified to include a reflection of the user that does not include the device that the user is holding or wearing. In another example, the CGR environment is modified so that virtual content, such as a newer version of the electronic device or a virtual wand, replaces the electronic device in the reflection. In another example, the CGR environment is modified so that virtual content, such as a user avatar, replaces the user in the reflection.

Abstract: A method of generating a segmentation confidence map by processing classification values each indicating a respective classification of a respective voxel of a retinal C-scan into a respective retinal layer class of a predefined set of retinal layer classes, the method comprising: generating, for each voxel, a respective confidence value which indicates a level of confidence in the classification of the voxel; for a retinal layer class of the predefined set, identifying a subset of the voxels such that the classification value generated for each voxel indicates a classification of the voxel into the retinal layer class; calculating, for each A-scan having voxels in the identified subset, a respective average of the confidence indicator values generated for the voxels; and using the calculated averages to generate the map, which indicates a spatial distribution of a level of confidence in the classification of the voxels.

Abstract: An object of the present disclosure is to reduce the number of pages while guaranteeing the consistency of the story of an album. One embodiment of the present invention is a non-transitory computer readable storage medium storing a program for causing a computer to function as: a selection unit configured to select, based on analysis results of each of a plurality of model images used in a model album and analysis results of each image in a candidate image group that is used in an album, similar images similar to the model images from among the candidate image group; a creation unit configured to create a layout in which the similar images selected by the selection unit are arranged in templates of the model album; and a reducing unit configured to reduce a number of pages of the layout.

Abstract: A process and an activation device (30) provides a user (B) activation of a drug measuring device (100). This drug measuring device includes an input unit (3), a testing unit (6) for testing given samples and an image recording device (12). The activation device (30) receives an activation image set (InB), which was generated by the drug measuring device or by a drug measuring device and at least one image (40, 40.1) from the image recording device. The received activation image set is checked as to meeting a predefined release criterion. For this the activation image set is compared with a predefined reference image set (35) for the user. If the release criterion is met, the drug measuring device is released for the user. An activation data set (32) for the user is completed by at least one image of the activation image set, which shows the user.

Abstract: Methods, systems and apparatuses may provide for technology that trains a neural network by inputting video data to the neural network, determining a boundary loss function for the neural network, and selecting weights for the neural network based at least in part on the boundary loss function, wherein the neural network outputs a pixel-level segmentation of one or more objects depicted in the video data. The technology may also operate the neural network by accepting video data and an initial feature set, conducting a tensor decomposition on the initial feature set to obtain a reduced feature set, and outputting a pixel-level segmentation of object(s) depicted in the video data based at least in part on the reduced feature set.

Abstract: A method, apparatus and storage medium for performing a video quality assessment (VQA) are provided. The method includes obtaining a plurality of images of a video, the plurality of images being divided into one or more groups; determining first images among the plurality of images to which a parametric-based VQA is to be applied, the first images being all of the plurality of images; determining a first score based on applying the parametric-based VQA to each of the first images; determining second images among the plurality of images to which a sample-based VQA is to be applied; determining a second score based on applying the sample-based VQA to each of the second images; and outputting a final score for at least one image based on the first score and the second score.

Abstract: A method includes: inputting an image of a target to each of a first model for estimating a region of an annular face of the target, a second model for estimating a region in an outer circumference of the annular face and a third model for estimating a region in an inner circumference of the annular face, and outputting a region image, as an estimated region of the annular face, the region image being obtained by synthesizing first, second, and third elements. The first element is a region outputted as a result of the first model for calculating a sum set in the synthesizing, the second element is a region outputted as a result of the second model for calculating a sum set in the synthesizing, and the third element is a region outputted as a result of the third model for calculating a difference set in the synthesizing.

Abstract: A method is for providing a second artifact-reduced x-ray image dataset based on an artifact-affected x-ray image dataset of an examination object, the artifact being caused by an object at least one of on, outside of and within the examination object. In an embodiment, the method includes creating a first artifact-reduced x-ray image dataset based on the artifact-affected x-ray image dataset, based on which a second projection dataset is created; identifying an object area which maps the object in the at least one projection; creating a third projection dataset based on the first projection dataset; and crating the second artifact-reduced x-ray image dataset based on the third projection dataset, through which the second artifact-reduced x-ray image dataset is provided.

Abstract: Provided is an image splicing method, including the following. A to-be-predicted image is divided into first and second cropped images. An overlap region exists between the first and second cropped images. The overlap region is divided into first and second sub-overlap regions. A first image region that does not include the second sub-overlap region is found in the first cropped image, and a second image region that does not include the first sub-overlap region is found in the second cropped image. First and second prediction result images respectively corresponding to the first and second cropped images are generated. A first prediction image region corresponding to the first image region is found in the first prediction result image, and a second prediction image region corresponding to the second image region is found in the second prediction result image. The first and second prediction image regions are spliced into a spliced image.

Abstract: Electronic devices, methods, and non-transitory program storage devices for leveraging machine learning to perform improved image fusion and/or noise reduction are disclosed. When a capture request is received, a neural network may be used to perform fusion and denoising operations on a first set of captured input images. According to some embodiments, the neural network's architecture comprises: a first plurality of network layers configured to compute optical flow information between the first set of input images; a second plurality of network layers configured to perform, at least in part, the fusion and denoising operations on the first set of input images; and a third plurality of skip connections between layers of the second plurality of network layers, wherein parameters for each skip connection of the third plurality of skip connections are warped based on at least part of the optical flow information computed by the first plurality of network layers.