Abstract: Techniques and constructs to facilitate automatic tagging can provide improvements in image storage and searching. The constructs may enable training a deep network using tagged source images and target images. The constructs may also train a top layer of the deep network using a personal photo ontology. The constructs also may select one or more concepts from the ontology for tagging personal digital images.

Abstract: A radiation imaging system including an X-ray sensor with a plurality of sensor chips disposed next to each other and an image processing apparatus which performs abnormality determination concerning a white image obtained by irradiating the X-ray sensor with X-rays without any object disposed between the X-ray sensor and an X-ray generator, the system includes: an area acquisition unit configured to acquire information of partial areas of the white image which respectively correspond to the plurality of sensor chips; a distribution acquisition unit configured to acquire distribution information representing variation in pixel value for each of the acquired partial areas; and a determination unit configured to determine, based on the distribution information, whether abnormality is included in the white image.

Abstract: Embodiments generally relate to providing systems and methods for assessing image quality of a distorted image relative to a reference image. In one embodiment, the system comprises a convolutional neural network that accepts as an input the distorted image and the reference image, and provides as an output a metric of image quality. In another embodiment, the method comprises inputting the distorted image and the reference image to a convolutional neural network configured to process the distorted image and the reference image and provide as an output a metric of image quality.

Abstract: Embodiments of the presently disclosed invention are directed to a document processing system and method that facilitates the processing and extraction of data from the documents. The system and method receive at least one document, where the document may contain data for extraction. The document may then be converted into a preferred document format and outputted to a user interface. The system and method may then receive a selection of at least a portion of the document, wherein the selection contains data for extraction. Based on the selection, at least one coordinate set corresponding to the selection and associated with at least one data field of interest is generated. Then the data from the selection of the document is extracted using the at least one coordinate set. Finally, a structured data set that includes the extracted data is generating, and storing in a computer memory.

Abstract: A method for identifying and enumerating candidate target cells within a biological fluid specimen is described. The method includes obtaining a biological fluid specimen, preparing the biological fluid specimen by staining cell features in the biological fluid specimen, capturing a digital image having a plurality of color channels of the biological fluid specimen, and applying image analysis to the digital image. A computer program product for identifying candidate target cells within a biological fluid specimen is also described. The computer program comprises instructions to cause a processor to carry out the image analysis.

Abstract: A method for image processing includes providing a microscopic image of a structure fabricated on a substrate and computer-aided design (CAD) data used in fabricating the structure. The microscopic image is processed by a computer so as to generate a first directionality map, which includes, for a matrix of points in the microscopic image, respective directionality vectors corresponding to magnitudes and directions of edges at the points irrespective of a sign of the magnitudes. The CAD data are processed by the computer so as to produce a simulated image based on the CAD data and to generate a second directionality map based on the simulated image. The first and second directionality maps are compared by the computer so as to register the microscopic image with the CAD data.

Abstract: In a mask inspection apparatus, a defect detection unit includes a first memory region, a second memory region, an inspection condition reconfiguring unit, and a comparison unit. The inspection condition reconfiguring unit obtains a difference between gray scales of an optical image stored in the second memory region and a reference image stored in the first memory region. The existence of the defect is determined in the case where the difference is larger than a first threshold value. Further in the case where the number of defects is larger than a second threshold value, an inspection condition is re-estimated and configured and the reference image is regenerated. A comparison unit compares the stored reference image with the stored optical image in the case where the inspection condition is not reconfigured, and compares the regenerated reference image with the stored optical image in the case where the inspection condition is reconfigured.

Abstract: The disclosed embodiments illustrate methods and systems for crowdsourcing a task. The method includes extracting one or more components from at least one task based on a task category associated with the at least one task. Each component corresponds to a part of the at least one task, that requires an input from a crowdworker. Further, the one or more components are categorized in a set of first categories, each of which includes similar components, and a second category that includes dissimilar components. The categorization is based on a comparison between content of each of component, and the task category. Thereafter, a representative component from at least one category from the set of first categories, and at least one component from the second category are combined to create the task. The task, so created, is crowdsourced to one or more crowdworkers.

Abstract: A method is provided that includes identifying a plurality of data sets, each data set is associated with a distribution model and each data set is associated with an image having a first noise level. The method includes partitioning the data sets into a plurality of groups and generating a best representative estimate for each group, the estimate is associated with a second noise level that is less than the first noise level. The method further includes annotating each group and receiving an input data set. The method includes assigning the input data set to a group and annotating the input data set according to that group's annotation.

Abstract: In real biometric systems, false match rates and false non-match rates of 0% do not exist. There is always some probability that a purported match is false, and that a genuine match is not identified. The performance of biometric systems is often expressed in part in terms of their false match rate and false non-match rate, with the equal error rate being when the two are equal. There is a tradeoff between the FMR and FNMR in biometric systems which can be adjusted by changing a matching threshold. This matching threshold can be automatically, dynamically and/or user adjusted so that a biometric system of interest can achieve a desired FMR and FNMR.

Abstract: A method, system, and computer product for detecting an oral temperature for a human include capturing a thermal image for a human using a camera, detecting a face region of the human from the thermal image, detecting a mouth region on the face region, comparing a temperature value on the mouth region to a reference temperature value on a first other face region, detecting an open mouth region on the mouth region based on a comparison result of the temperature value on the mouth region to the reference temperature value of the first other face region, determining whether a mouth of the human is open enough for an oral temperature to be detected, and computing the oral temperature based on temperature values on the mouth region and at least one other face region, responsive to the determination of the mouth being open enough for the oral temperature to be detected.

Abstract: Comprehensive 2D learning images are collected for learning subjects. Standardized 2D gallery images of many gallery subjects are collected, one per gallery subject. A 2D query image of a query subject is collected, of arbitrary viewing aspect, illumination, etc. 3D learning models, 3D gallery models, and a 3D query model are determined from the learning, gallery, and query images. A transform is determined for the selected learning model and each gallery model that yields or approximates the query image. The transform is at least partly 3D, such as 3D illumination transfer or 3D orientation alignment. The transform is applied to each gallery model so that the transformed gallery models more closely resemble the query model. 2D transformed gallery images are produced from the transformed gallery models, and are compared against the 2D query image to identify whether the query subject is also any of the gallery subjects.

Abstract: A method for capturing three-dimensional photographic lighting of a spherical lighting device is described. Calculation of boundaries of the spherical lighting device based on lighting properties of at least one light source in a set location of the spherical lighting device is performed. A mapping of multitude points of the spherical lighting device to three-dimensional vectors of at least one camera device using a logical grid is performed. A measurement of brightness of the logical grid of the spherical lighting device is performed. The method further comprises determining brightest grid point of the logical grid of the spherical lighting device, wherein the brightest grid point of the logical grid is measured within a region brightness of the spherical lighting device. The method further comprises calculating the region of brightness of the spherical lighting device based on the determined brightest grid point of the logical grid.

Abstract: A method, a device, and a non-transitory storage medium having instructions to invoke a setup of a fingerprint identification service of the user device, wherein the setup automatically sets up the fingerprint identification service without receiving a request from a user to set up the fingerprint identification service, capture, via the fingerprint reader, a fingerprint sample of the user of the user device; store the fingerprint sample; determine whether to generate a fingerprint template; generate the fingerprint template based on a determination to generate the fingerprint template; store the fingerprint template; provide a notification to the user that indicates that the fingerprint identification service can be activated; receive a request to activate the fingerprint identification service from the user; and activate the fingerprint identification service based on the request to activate the fingerprint identification service.

Abstract: An image is received that contains handwritten characters as answers to an inquiry. An automatic character recognition process is performed on the image to generate initially recognized characters from the handwritten characters. The initially recognized characters of an incorrect answer to a question of the inquiry are analyzed to automatically identify alternative recognized characters of the incorrect answer. Then, it can be determined whether one or more of the alternative recognized characters, when substituted in place of the initially recognized characters of the incorrect answer, change the incorrect answer to a correct answer to the question. If the alternative recognized characters substituted in place of the initially recognized characters in the correct answer exceed a minimum character recognition confidence value, the incorrect answer is changed to the correct answer to modify the initially scored answers to modified scored answers, and such are output.

Abstract: A dictionary for detection of an object is created from an image obtained by performing an image process, which depends on a first image process parameter, on the training image of the detection target object. The dictionary created based on the image process depending on the first image process parameter is determined, based on a result of detecting the object from an image obtained by performing an image process, which depends on the first image process parameter, on a photographed image based on the dictionary. A second image process parameter is determined, based on a result of detecting the object from an image obtained by performing an image process, which depends on the second image process parameter, on the photographed image using the determined dictionary.

Abstract: The present invention provides a method and an apparatus for detecting a salient region of an image. Classification processing is obtained by means of pre-training, so as to obtain a classification label, where the classification label is used to indicate a salience detection algorithm for detecting a salient region of the test image. Salience detection is performed on the test image by using the salience detection algorithm indicated by the classification label, so as to obtain the salient region of the test image. Because a salience detection algorithm with the best detection effect is acquired by using the image feature vector of the test image, to detect the salient region of the test image, accuracy of salience detection is improved.

Abstract: In real biometric systems, false match rates and false non-match rates of 0% do not exist. There is always some probability that a purported match is false, and that a genuine match is not identified. The performance of biometric systems is often expressed in part in terms of their false match rate and false non-match rate, with the equal error rate being when the two are equal. There is a tradeoff between the FMR and FNMR in biometric systems which can be adjusted by changing a matching threshold. This matching threshold can be automatically, dynamically and/or user adjusted so that a biometric system of interest can achieve a desired FMR and FNMR.

Abstract: This disclosure describes techniques for training models from video data and applying the learned models to identify desirable video data. Video data may be labeled to indicate a semantic category and/or a score indicative of desirability. The video data may be processed to extract low and high level features. A classifier and a scoring model may be trained based on the extracted features. The classifier may estimate a probability that the video data belongs to at least one of the categories in a set of semantic categories. The scoring model may determine a desirability score for the video data. New video data may be processed to extract low and high level features, and feature values may be determined based on the extracted features. The learned classifier and scoring model may be applied to the feature values to determine a desirability score associated with the new video data.

Abstract: A system and method of operation of an image processing system includes: a capture image sequence module for receiving an image sequence; a calculate iteration depth map module for calculating an iteration depth map and a zero-crossing array; a calculate median depth map module for calculating a median depth map and a depth difference map; a calculate variance depth map module for calculating a variance depth map; a calculate image depth map module for calculating an image depth map based on the iteration depth map, the zero-crossing array, the median depth map, the depth difference map, and the variance depth map; and a calculate display image module for calculating a display image based on the received images and the image depth map for displaying on a display device.