Abstract: The analyzing apparatus: generates first internal data; converts a position of first feature data in a feature space, based on the first internal data and a second learning parameter; reallocates, based on a result of first conversion and the first feature data, the first feature data to a position obtained through the conversion in the feature space; calculates a predicted value of a hazard function of analysis time in a case where the first feature data is given, based on a result of reallocation and a third learning parameter; optimizes the first to third learning parameters, based on a response variable and a first predicted value; generates second internal data, based on second feature data and the optimized first learning parameter; converts a position of the second feature data in the feature space, based on the second internal data and the optimized second learning parameter; and calculates importance data.

Abstract: Embodiments of the present disclosure describe a system and method for optical character recognition. In one embodiment, a system receives an image depicting text. The system extracts features from the image using a feature extractor. The system applies a first decoder to the features to generate a first intermediary output. The system applies a second decoder to the features to generate a second intermediary output, wherein the feature extractor is common to the first decoder and the second decoder. The system determines a first quality metric value for the first intermediary output and a second quality metric value for the second intermediary output based on a language model. Responsive to determining that the first quality metric value is greater than the second quality metric value, the system selects the first intermediary output to represent the text.

Abstract: The present invention relates to a computer-implemented method and a system for testing the output of a neural network (1) having a plurality of layers (11), which detects or classifies objects. The method comprises the step (S1) of reading at least one result from at least one first layer (11) and the confidence value thereof, which is generated in the first layer (11) of a neural network (1), and the step (S2) of checking a plausibility of the result by taking into consideration the confidence value thereof so as to conclude whether the object detection by the neural network (1) is correct or false. The step (S2) of checking comprises comparing the confidence value for the result with a predefined threshold value. In the event that it is concluded in the checking step (S2) that the object detection is false, output of the object falsely detected by the neural network is prevented.

Abstract: A method, system, and computer-readable media of generating a display on a device, including combining content from a plurality of sources into a display, the content from each of the plurality of sources being presented as a layer of the display, and further, each layer of the display being of substantially the same dimensions, detecting one or more objects in each layer of the generated display, and for one or more of the detected objects determining an object type or classification, determining if the object is overlapping or obscuring an object in a different layer of the generated display, and determining if the object will appear to a viewer as if it will overlap or obscure an object in a different layer of the generated display as a result of the motion, orientation, or gaze of the viewer.

Abstract: Described herein are systems, methods, and other techniques for training a generative adversarial network (GAN) to perform an image-to-image transformation for recognizing text. A pair of training images are provided to the GAN. The pair of training images include a training image containing a set of characters in handwritten form and a reference training image containing the set of characters in machine-recognizable form. The GAN includes a generator and a discriminator. The generated image is generated using the generator based on the training image. Update data is generated using the discriminator based on the generated image and the reference training image. The GAN is trained by modifying one or both of the generator and the discriminator using the update data.

Abstract: A method and device for automatically increasing the spectral bandwidth of an audio signal including generating a “mapping” (or “prediction”) matrix based on the analysis of a reference wideband signal and a reference narrowband signal, the mapping matrix being a transformation matrix to predict high frequency energy from a low frequency energy envelope, generating an energy envelope analysis of an input narrowband audio signal, generating a resynthesized noise signal by processing a random noise signal with the mapping matrix and the envelope analysis, high-pass filtering the resynthesized noise signal, and summing the high-pass filtered resynthesized noise signal with the original an input narrowband audio signal. Other embodiments are disclosed.

Abstract: A method for model customization according to an embodiment includes providing a user with prediction results of each of a plurality of pre-trained natural language processing models for a document subjected to analysis selected from a document set including a plurality of documents, acquiring user feedback on the prediction results from the user, generating a plurality of augmented documents from at least one of the plurality of documents based on data attributes of each of the plurality of documents and the user feedback; and retraining at least one of the plurality of natural language processing models, using training data including the plurality of augmented documents.

Abstract: An image tracking apparatus, which tracks a subject in a moving image, comprising: an image acquisition device that acquires image data indicating the moving image; a first detector that detects a first area of a subject in acquired image data to track the first area in the moving image; a second detector that detects a second area of a subject in the image data to track the second area in the moving image; and a controller that tracks the subject by switching between a tracking result of the first detector and a tracking result of the second detector. The first and the second detector operate independently. The controller tracks the subject using the tracking result of the first detector when the first area is tracked, and tracks the subject using the tracking result of the second detector when the first area is not tracked and the second area is tracked.

Abstract: The present disclosure relates to a dimension measuring device that shortens a time required for dimension measurement and eliminates errors caused by an operator.

Abstract: A computer-implemented method for determining field boundaries and crop forecasts in each field is provided. The method includes deriving vegetation indices for each geo-spatial pixel of each image of multi-spectral imagery at a plurality of points in time, constructing minimum bounding boxes for each image according to the vegetation indices, and generating, based on a neural network analysis of each image and the minimum bounding boxes, a geo-spatial plot of crops including a predicted plot of future crop usage for an area including each field in the multi-spectral imagery.

Abstract: In medical examination of breast cancer, a lesion computer-aided detection is performed in real time and with high accuracy, and a burden on a medical worker is reduced. A medical image processing apparatus that processes a medical image includes: a detection unit configured to detect a lesion candidate region; a validity evaluation unit configured to evaluate validity of the lesion candidate region by using a normal tissue region corresponding to the detected lesion candidate region; and a display unit configured to determine display content to a user by using an evaluation result.

Abstract: An image processing apparatus includes a detection unit that executes detection processing for detecting a particular object in an image, a holding unit that holds object information indicating a position and a size of the particular object on the image, a determination unit that determines whether a number of times a particular object is detected in the detection processing on one or more images reaches a predetermined value, a first setting unit that, when the number of times a particular object is detected in the detection processing on the one or more images is determined to reach the value, sets estimation areas on an image based on the object information obtained by the detection processing on the one or more images, and an estimation unit that executes estimation processing for estimating a number of the particular objects in the estimation areas.

Abstract: Systems, devices, media, and methods are presented for training a predictive model to detect objects in digital images, such as detecting whether a cotter key is present or absent in a digital photograph of a rail car coupler. The training system includes curating a plurality of training datasets, each including a number of raw images, together with a number of adjusted, augmented, and duplicate images. The predictive model includes a localization algorithm and an ensemble of models for classification. The localization algorithm is a deep convolutional neural network (CNN) which identifies a region of interest. One of more of the deep CNN classification models generates a plurality of candidate regions associated with each region of interest, thereby generating a large number of additional regions useful for training. In use, the trained predictive model is part of a detection and notification system that processes new images from the field and broadcasts a notice when an anomaly is detected.

Abstract: An operating method with goods information is applicable to an electronic device. The operation method includes: obtaining image information associated with one or more goods objects on a target electronic shelf among a plurality of electronic shelves in a network; and performing first communicating with a server for controlling the electronic shelves in the network according to either or both of the image information and feature information associated with the one or more goods objects, wherein feature information is extracted from the image information and the first communicating includes wirelessly transmitting either or both of the image information and the feature information associated with the one or more goods objects to the server.

Abstract: A system, method, and computer program product for automatically identifying a casualty and matching an electronic health record (EHR) to the casualty. A casualty identification is determined by matching a presenting image of the casualty with one of a master image or a social media profile image of the casualty. A recognized patient messaging system (RPMS) is configured to query one or more electronic health records (EHR) service providers for the existence of an EHR corresponding the identified casualty and automatically communicate the existence of the EHR to the EMS provider. A blockchain trusted identification module (BTIM) is configured to establish a trust relationship between the EMS provider and the one or more EHR service providers to establish a trusted pathway for delivery of the casualty's EHR to the EMS provider. With the casualty's EHR emergency responders to can provide better care for the casualty in an emergency situation.

Abstract: A method for distinguishing between types of intraoral areas of interest (AOIs) includes receiving or generating a first 3D model comprising teeth; receiving or generating a second 3D model comprising the teeth; comparing the first 3D model to the second 3D model; determining a difference between the first 3D model and the second 3D model at a dental site comprising a tooth; determining an intraoral AOI that is associated with the difference; determining whether the difference comprises a first type of change indicative of a first type of intraoral AOI or a second type of change indicative of a second type of intraoral AOI; and responsive to determining that the difference comprises the first type of change, determining that the intraoral AOI is the first type of intraoral AOI and generating an indicator of the first type of intraoral AOI for the tooth.

Abstract: A position information estimating method and apparatus are provided, wherein the position information estimating method includes estimating initial position information of a vehicle based on motion information of the vehicle that is obtained from one or more sensors, calculating search position information of one or more objects included in an image of surroundings of the vehicle based on the initial position information, evaluating a final reliability of each of the one or more objects based on the initial position information and the search position information, obtaining a filter adaptation parameter corresponding to a weight of the search position information based on the final reliability, and estimating final position information of the vehicle based on the initial position information, the search position information, and the filter adaptation parameter.

Abstract: The disclosure discloses a power transformer winding fault positioning method based on deep convolutional neural network integrated with visual identification, including 1) a winding equivalent circuit is established, and a transfer function thereof is calculated; 2) a sine wave excitation source is set at one end of the power transformer winding to obtain the amplitude-frequency characteristic curve of each winding node; 3) circuits under various fault statuses are subjected to scanning frequency response analysis to extract amplitude-frequency characteristics; 4) a feature matrix is established based on the obtained amplitude-frequency characteristics; 5) scanning frequency response analysis is performed on the diagnosed power transformer to form a feature matrix; 6) the feature matrix is converted into an image, simulation and historical detection data are used as a training set, and a deep convolutional neural network is input for training; 7) diagnosed transformer is subjected to fault classification and

Abstract: A method of analyzing a substrate contacting a fluid present in an industrial system is provided. The method comprises creating a series of digital images of the substrate while contacting the fluid present in the industrial system. A region of interest in the series of digital images of the substrate is defined. A corrosion feature in the region of interest in the series of digital images of the substrate is identified. The corrosion feature in the region of interest in the series of digital images of the substrate is analyzed to determine a corrosion trend of the industrial system. In certain embodiments of the method, the fluid is industrial water, and the industrial system is an industrial water system.

Abstract: Systems for and methods of monitoring and analyzing deposit in an industrial water system are provided. The methods comprise heating a substrate while the substrate contacts industrial water in the industrial water system to form deposit on the substrate. A series of digital images of the substrate while the substrate contacts the industrial water in the industrial water system is created. A region of interest in the series of digital images of the substrate is defined. A deposit feature in the region of interest in the series of digital images of the substrate is identified. The deposit feature in the region of interest in the series of digital images of the substrate is analyzed to determine a deposit trend of the substrate in the industrial water system. Generally, the systems are configured so as to be capable of carrying out one or more of the methods.