Abstract: Embodiments described herein include, software, firmware, and hardware logic that provides techniques to perform arithmetic on sparse data via a systolic processing unit. Embodiment described herein provided techniques to detect zero value elements within a vector or a set of packed data elements output by a processing resource and generate metadata to indicate a location of the zero value elements within the plurality of data elements.

Abstract: A method and system for quantifying semantic variance between neural network representations is provided. Two neural network representations to be compared are first extracted, the weight of each filter in an intermediate layer corresponding to each semantic concept is learned on a reference dataset using the Net2Vec method, then set IoU of each representation for all semantic concepts in the reference dataset are calculated, and finally variance between the set IoU of the two representations for all the semantic concepts are integrated to obtain semantic variance between the two neural network representations. The method solves the problem of lack of accurate measurement on the variance between neural network representations on a semantic information level, and has an accurate measurement effect.

Abstract: Provided are an augmented handwritten signature authentication method and an electronic device supporting the same, the method including: receiving a primary touch input corresponding to a primary handwritten signature input of a user; receiving, after the primary handwritten signature input is received and a specified time period elapses, a secondary mark input of the user; obtaining primary touch position data corresponding to the primary touch input, and the secondary mark input; and performing user authentication by comparing the primary touch position data corresponding to the primary touch input and secondary touch position data related to at least part of the secondary mark input with prestored primary touch registration data and secondary touch registration data, respectively.

Abstract: A method including receiving a first set of one or more street level images of houses into a Machine Learned Model trained with a second set of street level images with one or more exterior features of the houses labeled, identifying the one or more exterior features in the first set of street level images by way of the Machine Learned Model, and quantifying and outputting the counts and/or two-dimensional areas for each of the identified exterior features in the first set of one or more street level images is described. Non-transitory, computer-readable storage media having instructions for executing the method steps by one or more processors as well as computer or computer systems capable of performing the method steps are also described.

Abstract: A method including receiving a first set of one or more street level images of houses into a Machine Learned Model trained with a second set of street level images with one or more exterior features of the houses labeled, identifying the one or more exterior features in the first set of street level images by way of the Machine Learned Model, and quantifying and outputting the counts and/or two-dimensional areas for each of the identified exterior features in the first set of one or more street level images is described. Non-transitory, computer-readable storage media having instructions for executing the method steps by one or more processors as well as computer or computer systems capable of performing the method steps are also described.

Abstract: A system includes first and second imaging assemblies, and a processor. The first imaging assembly is configured to produce a first image of a measurement site in a sample. The second imaging assembly is coupled with a measurement assembly and is configured to produce a second image of the measurement site. The processor is configured to: (i) perform, based on the first image, a first movement of the sample relative to the measurement assembly, (ii) perform, based on the second image, a second movement of the sample for aligning the sample with the measurement assembly, and (iii) control the measurement assembly to perform a measurement in the measurement site.

Abstract: A gaze estimation apparatus according to one or more embodiments may estimate the gaze direction of a target person using calibration information including feature information about the gaze of the eyes of the target person looking in a predetermined direction and true value information indicating the true value for the predetermined direction, in addition to a target image including the eye(s) of the target person. The gaze estimation apparatus may thus estimate the gaze direction of the target person reflecting individual differences, and may allow estimation of the gaze direction of the target person with higher accuracy.

Abstract: In a method for training a neural network to recognize a tool condition based on image data, the neural network is trained to recognize the tool condition of a first tool type, and image data of a second tool type is applied. The image data is subjected to image processing. Via this, the image data of the second tool type is converted into image data of the first tool type. The neural network is trained based on the converted image data. In a method for machining and/or production via the first tool type, the tool condition of the first tool type is recognized via a neural network that is trained in accordance with such a method.

Abstract: A data input device includes: a controller that: acquires position data indicating position coordinates of a facility, estimates, based on the position data, one or more facility candidates, and causes a display to display an input screen showing the estimated one or more facility candidates as options for selection by a user, and creates inspection data relating to gas inspection of the one or more facility candidates selected by the user.

Abstract: A data encoding method includes determining a value of a syntax element of point cloud data of a point cloud. The point cloud data includes attribute values of the point cloud and the syntax element indicates a search range of the attribute values during prediction encoding. The method further includes performing prediction processing on the attribute values according to the search range to obtain residuals of the attribute values, and encoding the residuals and a difference between the value of the syntax element and a constant value to generate code stream data. The constant value is a positive number.

Abstract: To, in a case where editing of a common double-page spread is performed on one user's album, selectively apply the contents of that editing result on a per user basis, a program is provided which causes a computer to function as: an editing unit configured to edit layouts of albums each having an attribute of a common double-page spread common to all albums and an attribute of an individual double-page spread in each album; a first determination unit configured to determine whether an edited double-page spread in a predetermined album on which editing is performed is a common double-page spread; and a second determination unit configured to determine, for each album other than the predetermined album, whether to apply a content of a result of the editing of the predetermined album to the album in a case where a result of the determination by the first determination unit is positive.

Abstract: An image processing method, applied to an electronic apparatus, includes performing feature extraction processing on a source image to obtain a feature data set corresponding to a source file, determining an image processing parameter corresponding to the feature data set according to the feature data set, and processing the source image according to the image processing parameter.

Abstract: A method for improving the reconstruction of images obtains an object image and a reference image and extracts a first edge of the object image and a second edge of the reference image based on a predetermined algorithm. A first vector of the plurality of first pixels to the first edge and a second vector of the plurality of second pixels to the second edge are obtained and a determination made as to whether the first vector and the second vector are consistent. A loss between the first vector and the second vector is calculated if the first and second vector are not consistent and a predetermined model is corrected based on the loss, the reference image being reconstructed into the object image based on corrected predetermined model. An image reconstruction device and a non-transitory storage medium are also disclosed.

Abstract: Methods for business operation based on vehicle-carrying image processing by using an electronic computer device are presented, the methods include: causing the electronic computing device to capture, transmit, model and interpret the images; causing the business to operate based on the assembled information conveyed through the images; causing the automatic establishment of bid processes for business transactions. In some embodiments, methods further include: mapping the Quick Response (QR) codes to the classified objects; building and applying deep learning models for pattern recognition; building and applying an embedded communication network. In some embodiments, methods further include: determining the vehicle operation status and operation irregularities. In some embodiments, methods further include: causing the searches on the embedded communication network to generate inquiries and to establish bid processes.

Abstract: Described herein are systems, methods, and non-transitory computer readable media for validating or rejecting automated detections of an entity being tracked within an environment in order to generate a track representative of a travel path of the entity within the environment. The automated detections of the entity may be generated by an artificial intelligence (AI) algorithm. The track may represent a travel path of the tracked entity across a set of image frames. The track may contain one or more tracklets, where each tracklet includes a set of validated detections of the entity across a subset of the set of image frames and excludes any rejected detections of the entity. Each tracklet may also contain one or more user-provided detections in scenarios in which the tracked entity is observed or otherwise known to be present in an image frame but automated detection of the entity did not occur.

Abstract: A system for monitoring vehicle traffic includes one or more non-transitory computer-readable storage media storing computer-executable instructions that, when executed by one or more processors, cause a computing system to recognize, using images, at least one of a first face of a person in a facial detection zone, compare the first face to a database, identify the person as a suspicious person, and correlate the first face with a stored electronic device identifier. The computer-executable instructions further cause the computing system to access at least one personal identification database, and retrieve personal identification information of a suspicious individual by correlating information of the personal identification database with the first face or the stored electronic device identifier.

Abstract: A reduced noise image can be formed from a set of images. One of the images of the set can be selected to be a reference image and other images of the set are transformed such that they are better aligned with the reference image. A measure of the alignment of each image with the reference image is determined. At least some of the transformed images can then be combined using weights which depend on the alignment of the transformed image with the reference image to thereby form the reduced noise image. By weighting the images according to their alignment with the reference image the effects of misalignment between the images in the combined image are reduced. Furthermore, motion correction may be applied to the reduced noise image.

Abstract: A computer hardware system having an image modification platform includes a hardware processor configured to initiate the following executable operations. A digital image is received by the image modification platform. A plurality of objects within the digital image are detected using an object segmentation engine of the image modification platform. The plurality of objects are classified using an object classification engine of the image modification platform and involve associating a stakeholder, respectively, to at least two of the plurality of objects. A knowledge map for the digital image is generated based upon the classifying and the stakeholders using a knowledge map engine. A set of constraints are identified for each of the plurality of objects based upon the knowledge map and the plurality of stakeholders. A modified digital image to be published is generated by an interactive image generation engine using the sets of constraints.

Abstract: A synesthesia-based encryption system and method (referred to as a system) includes a camera that captures an image and a transceiver communicatively coupled to the camera and a video-only network. The system includes a sensor that monitors a location and generates a sensor message. The sensor message include information that represents a state, a measurement, and/or a detection at that location. The system's processor maps colors to characters from the sensor message to generate a replacement image. In some systems, the sensor is encrypted first. The processor integrates the replacement image within the original image or some or all of the video frames captured by the camera to form a combined image(s) and causes a transceiver to transmit the combined image(s) across the video-only network to a destination.

Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed to decode receipts based on neural graph architecture. An apparatus includes interface circuitry to obtain an image of a document; machine readable instructions; and programmable circuitry to execute the machine readable instructions to at least generate nodes for a feature graph based on features extracted from text boxes, the nodes including polar coordinates indicative of angular positions of the text boxes; pass the feature graph through a graph neural network to generate an adjacency matrix; and identify text lines in the image of the document by clustering the nodes by line based on the adjacency matrix.