Abstract: A method, computer system, and a computer program product for predicting a variation of sequential blood glucose levels by using deep learning is provided. The present invention may include training a predictor associated with a user by using a deep learning network. The present invention may further include predicting a plurality of sequential blood glucose levels by the trained predictor based on at least one meal image, at least one time-period, and at least one set of data associated with a plurality of blood glucose levels of the user.

Abstract: A text region positioning method and device, and a computer readable storage medium, which relate to the field of image processing. The text region positioning method includes acquiring a variance graph on the basis of an original image; acquiring an edge image of the variance graph; if a difference value among distances between edge points of opposing positions in two adjacent edge lines in the edge image is within a preset distance difference range, then the region between the two adjacent edge lines is determined as a text region.

Abstract: This disclosure describes techniques for identifying users that are enrolled for use of a user-recognition system and updating enrollment data of these users over time. To enroll in the user-recognition system, the user may initially scan his or her palm. The resulting image data may later be used when the user requests to be identified by the system by again scanning his or her palm. However, because the characteristics of user palms may change over the time, the user-recognition system may continue to build more and more data for use in recognizing the user, in addition to removing older data that may no longer accurately represent current characteristics of respective user palms.

Abstract: A method is presented for determining a weight of all of the grains of an ear of a grain stalk. The method comprises providing a digital image of the ear in a spindle step view of the ear, wherein the ear is located in front of a reference card as a background, and determining the length of the ear by means of a comparison with the reference card. Furthermore, the method comprises determining a number of spindle steps of the ear by means of a template matching method, determining a number of grains of the ear and determining of the weight of all of the grains of the ear.

Abstract: An image is printed by using a first, second and third ink. A surface tension difference between a first ink and a third ink is smaller than a surface tension difference between the first ink and a second ink. A first gradation value for the first ink, a second gradation value for the second ink and a third gradation value for the third ink are quantized to generate a first quantized value, a second quantized value, and a third quantized value respectively. This quantization processing is performed such that, the number of pixels for each of which the first quantized value indicates printing, a the second quantized value indicates non-printing and the third quantized value indicates non-printing is greater than the number of pixels for each of which the first quantized value indicates printing and at least the second quantized value indicates printing or the third quantized value indicates printing.

Abstract: Methods and apparatus, including computer program products, are provided for watermarking neural networks. In some embodiments, there may be provided a method. The method may include determining, for a neural network, an activation layer output by a hidden layer of the neural network. The method may include selecting a watermarking process. The method may include applying the selected watermarking process to the activation layer output to generate a key. The method may include storing, for the neural network to enable detection of copying of the neural network, the selected watermarking process and the key. Related systems, methods, and articles of manufacture are also described.

Abstract: An inspection device includes: an image acquirer that acquires an inspection target image; an edge extractor that extracts an edge from each of the inspection target image and a reference image to be used in inspecting the output image; a defect candidate region extractor that extracts a defect candidate region having a possibility of a defect by comparing the inspection target image with the reference image; an edge direction calculator that calculates a direction of the edge in the inspection target image and a direction of the edge in the reference image; and a defect determiner that determines whether the defect candidate region is a defect, on a basis of the direction of the edge in the inspection target image and the direction of the edge in the reference image at a position corresponding to the defect candidate region.

Abstract: The occurrence of more and the deterioration of performance are suppressed. A data compression unit stores identification data indicating the presence or absence of periodicity, and printing data correlated with the identification data in a first memory area of a RAM. A color conversion processing unit that performs speculation processing stores printing data that has undergone color conversion processing corresponding to printing by a printer in a second memory area of the RAM. A screen processing unit that performs speculation processing performs AM screen processing on printing data. The screen processing unit confirms the identification data stored in the first memory area of the RAM by the data compression unit, and in a case where the identification data indicates there is periodicity, instructs to perform color conversion processing corresponding to printing by a copier on printing data, and performs FM screen processing on the printing data.

Abstract: An information processing apparatus for managing a printing apparatus in a print site including at least one printing apparatus includes a memory that stores a set of instructions, and at least one processor that executes the instruction to determine a custom color that is used in a color verification operation of the at least one printing apparatus, obtain an input profile of a color verification setting of the determined custom color, display a screen for selecting whether or not to add a standard color corresponding to the obtained input profile to the color verification operation, and according to having received an instruction, via the screen, for adding the standard color to the color verification operation, instruct the at least one printing apparatus to perform printing in accordance with the color verification setting, including the input profile, for the custom color and the standard color.

Abstract: An example is provided to calculate calibration data for or print quality data of a printer, in one example, image data on printed elements of a plurality of patches printed at different powers or intensities by a print device may be determined.

Abstract: The inventive plastic card permits a scuba diver to photograph underwater such as at 100 feet, without using external lighting, and to achieve quality photographic images as if external lighting actually had been used during the underwater photography.

Abstract: In some implementations a neural network is trained to perform to directly predict thin boundaries of objects in images based on image characteristics. A neural network can be trained to predict thin boundaries of objects without requiring subsequent computations to reduce the thickness of the boundary prediction. Instead, the network is trained to make the predicted boundaries thin by effectively suppressing non-maximum values in normal directions along what might otherwise be a thick predicted boundary. To do so, the neural network can be trained to determine normal directions and suppress non-maximum values based on those determined normal directions.

Abstract: A vehicle-mounted image processing device 1000 acquires an image around a host vehicle 10, senses a feature amount of a target object present on a road surface from the image, and relative position coordinates with respect to the host vehicle 10, acquires information of a three-dimensional object around the host vehicle 10 and relative position coordinates of the three-dimensional object, distinguishes whether the feature amount is a feature amount of a feature on the road surface or a feature amount of a feature on the three-dimensional object by using a positional relationship between the feature amount and the three-dimensional object, and recognizes the target object by using a feature on the road surface. As a result, a vehicle-mounted image processing device, which is capable of avoiding misrecognition of a part of a three-dimensional object as a target object when sensing the target object present on a road surface such as a parking bay by using a camera image, is provided.

Abstract: Methods, systems, and non-transitory computer readable medium are described for automated image measurement for process development and optimization. A method includes receiving an image of a product associated with a manufacturing process; determining, using a trained machine learning model, an image classification for the image; selecting, based on the image classification, one or more image processing algorithms for the image; pre-processing the image based on at least one of the one or more image processing algorithms to generate an enhanced image; measuring, using a first image processing algorithm of the one or more image processing algorithms, one or more attributes of the enhanced image to determine image measurements; and reporting the image measurements. The manufacturing parameters of the manufacturing process are to be updated based on the image measurements.

Abstract: An information processing apparatus capable of providing an appropriate IP address to a mobile terminal in which a network service discovery protocol does not function. The information processing apparatus has a plurality of lines assigned with different pieces of address information, and is configured to transmit short-range wireless-communication information, in which address information regarding one line out of the plurality of lines is set, to a communication destination via short-range wireless communication. The information processing apparatus comprises a setting unit configured to preferentially set, into the short-range wireless-communication information, address information regarding a line, out of the plurality of lines, in which a relay device for performing intercommunication between different protocols is provided between the information processing apparatus and the communication destination.

Abstract: An image processing apparatus that converts image data of a first resolution into image data of a second resolution lower than the first resolution. The image processing apparatus determines a small dot that is included in the image data of the first resolution and matches a first small dot pattern and performs conversion from the first resolution to the second resolution by switching a method of generating resolution-converted image data based on a result of the determination.

Abstract: According to one embodiment, an image processing apparatus includes a scanner to generate image data by scanning a document and a control unit. The control is configured to set, for each pixel in the image data, a second pixel value based on a first pixel value and a difference between a first base value corresponding to a base color of the document and a second base value that is based on neighboring pixel values around the pixel. A correction table is used to obtain a corrected second pixel value for each pixel. A third pixel value for each pixel is set based on the corrected second pixel value and the difference between the first and second base values.

Abstract: Embodiments described herein provide a two-stage online detection of action start system including a classification module and a localization module. The classification module generates a set of action scores corresponding to a first video frame from the video, based on the first video frame and video frames before the first video frames in the video. Each action score indicating a respective probability that the first video frame contains a respective action class. The localization module is coupled to the classification module for receiving the set of action scores from the classification module and generating an action-agnostic start probability that the first video frame contains an action start.

Abstract: An image recognition device sets an overall observation region which surrounds a whole body of an object and partial observation regions which surround characteristic parts of the object respectively to locations in an image which are estimated to include captured images of the object. The device clips images in the overall observation region and the partial observation regions, and calculates similarity degrees between them and previously learned images on the basis of a combination of two image feature amounts. The device calculates an optimum ratio in combining the HOG feature amount and the color distribution feature amount individually for the regions. This ratio is determined by setting a weight parameter ?i for setting a weight used for combining the HOG feature amount and the color distribution feature amount to be included in a state vector and subjecting the result to complete search by a particle filter.

Abstract: A machine learning based image processing architecture and associated applications are disclosed herein. In some embodiments, a machine learning framework is trained to learn low level image attributes such as object/scene types, geometries, placements, materials and textures, camera characteristics, lighting characteristics, contrast, noise statistics, etc. Thereafter, the machine learning framework may be employed to detect such attributes in other images and process the images at the attribute level.