Abstract: Disclosed are various embodiments for processing multiple documents in an image. First, text from each of one or more documents in an image can be identified. An orientation of each of the one or more documents can be determined based at least in part on an alignment of the text of each of the one or more documents. Additionally, using an object detection model, the one or more documents in the image can be identified based at least in part on the orientation of each of the one or more documents. Finally, the one or more documents can be separated from the image into one or more separate image files, each separate image file representing a respective document of the one or more documents.

Abstract: A system described herein may provide a technique for using modeling techniques to identify events, trends, etc. in a set of data, such as streaming video or audio content. The system may perform lightweight pre-processing operations on a different set of data, such as object position data, to identify timeframes at which an event may potentially have occurred, and the modeling techniques may be performed at portions of the streaming content that correspond to such timeframes. The system may forgo performing such modeling techniques at other portions of the streaming content, thus conserving processing resources.

Abstract: The present invention provides a processing apparatus (10) including an acquisition unit (11) that acquires an image of a fill-in form including a plurality of first fill-in fields where a numerical value is filled in, and a second fill-in field where a sum total of the numerical values filled in a plurality of the first fill-in fields is filled in, an analysis unit (12) that analyzes the image, and recognizes the value filled in a plurality of the first fill-in fields and the value filled in the second fill-in field, a determination unit (13) that determines whether a sum total of recognition results of the value filled in a plurality of the first fill-in fields and a recognition result of the value filled in the second fill-in field match each other, and a processing unit (14) that executes error processing when a sum total of the recognition results of the value filled in a plurality of the first fill-in fields and the recognition result of the value filled in the second fill-in field do not match each oth

Abstract: In some aspects, the disclosure is directed to methods and systems for detection and classification of stamps in documents. The system can receive image data and textual data of a document. The system can pre-process and filter that data, and covert the textual data to a term frequency inverse document frequency (TF-IDF) vector. The system can detect the presence of a stamp on the document. The system can extract a subset of the image data including the stamp. The system can extract text from the subset of the image data. The system can classify the stamp using the extracted text, the image data, and the TF-IDF vector. The system can store the classification in a database.

Abstract: Methods, systems, and computer-readable storage media for determining that a subject is a live person include capturing a set of images of a subject instructed to perform a facial expression. A region of interest for the facial expression is determined in a first image of the set, the first image representing a first facial state that includes the facial expression. A set of facial features is identified in the region of interest, the facial features being indicative of interaction between facial muscles and skin of the subject due to the subject performing the facial expression. A determination is made, based on the facial features, that the first image substantially matches a template image of the facial expression of the subject. Responsive to determining that the first image substantially matches the template image, identifying the subject as a live person.

Abstract: An image drawing method, a display apparatus, and a storage medium, relates to the technical field of display. The method comprises: determining a gray-scale value of each of a plurality of target-pixels in a displayed image; determining a pixel quantity in each preset gray-scale level according to the gray-scale value of each of the plurality of target-pixels; representing a maximum quantity of a plurality of pixel quantities by a first numeral system, to read values in a preset digit range of the digit of the first numeral system corresponding to the maximum quantity, determining a proportional coefficient corresponding to the values, and according to a height of a histogram to be drawn, a gray-scale to be drawn and the proportion coefficient, drawing the histogram.

Abstract: The present disclosure is directed to systems and methods that enable scanning of any type of card regardless of the shape and design of a given card and/or a font, a shape and a format with which characters such as numbers, letters and symbols are printed on the cards including cards with non-embossed characters printed thereon. In one example, a method includes scanning a card, the card including at least an account number associated with a user of the card and an identifier of the user; detecting, by applying a machine learning model to the card after scanning the card, at least the account number printed on the card; and completing a task using the account number.

Abstract: A computing system engages in digital image processing of received video frames to generate sport data that indicates a score and/or a time associated with a sport event. The digital image processing includes: (i) identifying a first frame region of the video frames based on the first frame region depicting a scoreboard; (ii) executing a first procedure that analyzes the identified first frame region to detect, within the identified first frame region, second frame region(s) based on the second frame region(s) depicting text of the scoreboard; (iii) in response to detecting the second frame region(s), executing a second procedure to recognize the text in at least one of the second frame region(s); and (iv) based at least on the recognizing of the text, generating the sport data. In response to completing the digital image processing, the computing system then carries out an action based on the generated sport data.

Abstract: An approximate camera location on a route can be determined by inputting a first image acquired by a vehicle camera to a first convolutional neural network. First image feature points can be extracted from the first image using a feature extraction algorithm. Pose estimation parameters for a pose estimation algorithm can be selected based on the approximate camera location. A six degree-of-freedom (DoF) camera pose can be determined by inputting the first image feature points and second feature points included in a structure-from-motion (SfM) map based on the route to the pose estimation algorithm which is controlled by the pose estimation parameters. A six DoF vehicle pose can be determined based on the six DoF camera pose.

Abstract: This application relates to an image data inspection method and apparatus in the field of artificial intelligence (AI) technologies. The method includes obtaining an image to be inspected, the image to be inspected comprising a sequence of slice images; determining a corresponding group of slice images for each target image in the sequence of slice images; extracting a corresponding slice feature map for each slice image in the group of slice images; aligning the slice feature maps extracted corresponding to the group of slice images; aggregating context information of each slice image in the group of slice images by using an aligned feature map; and performing target region inspection on an aggregated feature map, to obtain an inspection result corresponding to the target image, and combining the inspection result corresponding to each target image, to generate an inspection result corresponding to the image to be inspected.

Abstract: Methods and systems that provide data privacy for implementing a neural network-based inference are described. A method includes injecting stochasticity into the data to produce perturbed data, wherein the injected stochasticity satisfies an ?-differential privacy criterion and transmitting the perturbed data to a neural network or to a partition of the neural network for inference.

Abstract: A method for preparing digital image data from an analog image input by scanning, and reducing visibility of the scanning noise, may include estimating a visibility of scanning noise, and a number of scanning samples needed to reduce scanning noise to below a visible threshold. Related methods include scanning, by an analog-to-digital image scanner, an analog image for multiple iterations, resulting in digital image data for each of the iterations; calculating a noise statistic for individual pixels of digital image data across the iterations; determining true values of individual pixels of the digital image data based on the noise statistic for each of the individual pixels and generating scanner noise reduced digital image data wherein pixels are assigned their respective ones of the true values; and saving the scanner noise reduced digital image data in a computer memory.

Abstract: The present disclosure relates to normalizing pixel intensity values of images. Qualitative images which have to be compared need to have actual pixel intensity values in a suitable range. Generally, the qualitative images to be compared are made to appear visually similar by adjusting display settings. First and second images having different actual pixel intensity values but appearing visually same are received. One or more normalized parameters are determined and using one or more contrast and illumination parameters associated with the first and second image. Thereafter, normalized pixel intensity values are determined using the one or more normalized parameters and a normalized image is generated. The pixel intensity values of the normalized image can be compared with actual pixel intensity values of the first image or second image.

Abstract: Biometric gallery management is performed by association one or more wireless identifiers that correspond to one or more mobile devices (such as smart phones, tablet computing devices, cellular telephones, wearable devices, smart watches, fitness monitors, digital media players, medical devices, and/or other mobile computing devices) that people carry with digital representations of biometrics corresponding to the people. Wireless identifiers corresponding to mobile devices proximate to a biometric reader device may be monitored. Upon detection of wireless identifiers corresponding to mobile devices proximate to the biometric reader device, the associated digital representations of biometrics may be loaded from a main gallery into one or more local galleries, which may then be used to perform one or more biometric identifications and/or verifications.

Abstract: Systems and methods for identifying and segmenting objects from images include a preprocessing module configured to adjust a size of a source image; a region-proposal module configured to propose one or more regions of interest in the size-adjusted source image; and a prediction module configured to predict a classification, bounding box coordinates, and mask. Such systems and methods may utilize end-to-end training of the modules using adversarial loss, facilitating the use of a small training set, and can be configured to process historical documents, such as large images comprising text. The preprocessing module within the systems and methods can utilize a conventional image scaler in tandem with a custom image scaler to provide a resized image suitable for GPU processing, and the region-proposal module can utilize a region-proposal network from a single-stage detection model in tandem with a two-stage detection model paradigm to capture substantially all particles in an image.

Abstract: An apparatus including a processor caused to receive document images, each including representations of characters. The processor is caused to parse each document image to extract, based on structure type, subsets of characters, to generate a text encoding for that document image. For each document, the processor is caused to extract visual features to generate a visual encoding for that document image, each visual feature associated with a subset of characters. The processor is caused to generate parsed documents, each parsed document uniquely associated with a document image and based on the text and visual encoding for that document image. For each parsed document, the processor is caused to identify sections uniquely associated with section type. The processor is caused to train machine learning models, each machine learning model associated with one section type and trained using a portion of each parsed document associated with that section type.

Abstract: Broadly speaking, the present techniques relate to a method, apparatus and system for improving a user's engagement or emotion during a human-computer interaction, by dynamically adjusting the human-computer interaction in response to detected user emotion.

Abstract: A system includes a computing platform having a hardware processor and a memory storing a software code and a neural network (NN) having multiple layers including a last activation layer and a loss layer. The hardware processor executes the software code to identify different combinations of layers for testing the NN, each combination including candidate function(s) for the last activation layer and candidate function(s) for the loss layer. For each different combination, the software code configures the NN based on the combination, inputs, into the configured NN, a training dataset including multiple data objects, receives, from the configured NN, a classification of the data objects, and generates a performance assessment for the combination based on the classification. The software code determines a preferred combination of layers for the NN including selected candidate functions for the last activation layer and the loss layer, based on a comparison of the performance assessments.

Abstract: A camera system may automatically identify high frequency motion regions in a field of view of the camera system. Analysis of motion events detected at different regions of a scene in the field of view, may be performed. Similar events frequently occurring at a certain region may be recorded/accumulated and compared to a threshold. Regions with high frequency activity may be determined and a user may be notified. If the user is not interested in events within the high activity regions, the regions may be ignored and notifications to the user may cease. The camera system may determine specific actions/events occurring in a region, such as a person approaching or leaving a home.

Abstract: A receipt capture tool residing on a customer mobile device may be initiated when a customer completes an in-store or online purchase. The receipt capture tool may prompt the customer to capture an image of a receipt detailing a purchase and an item (e.g., product or service) purchased. For instance, the photo of a physical receipt may be taken by the mobile device, or an electronic receipt or email detailing the purchasing transmitted from a physical merchant or online merchant server may be stored. Receipt information may be extracted and saved with other information pertinent to the item purchased, including warranty information. If the customer needs to return or repair the item purchased at a future date, the receipt and warranty information may be subsequently accessed via their mobile device. The receipt and warranty information may also be stored in a searchable database to facilitate easy retrieval by the customer.