Automatic document classification using machine learning may involve receiving inputs that assign documents to classifiers, which define document classification rules for a classification model. The computing device may train the classification model using a machine learning technique that assigns each document of a second set of documents to destinations based on the document classification rules. The computing device may also receive a template design for each destination that specifies metadata to extract for a document type corresponding to documents assigned to the destination. The computing device may subsequently classifying a particular document using the classification model, which may involve assigning the particular document to a given destination of the plurality of destinations based on the document classification rules, and exporting metadata from the particular document using the template design associated with the given destination.

An apparatus for automatic target recognition with reinforcement learning is provided. The apparatus receives an image of a scene and performs an automatic target recognition on the image to detect objects in the image as candidate targets. The apparatus divides the candidate targets into subsets of candidate targets and performs a verification of the automatic target recognition to identify true targets in the image. In the verification, the apparatus solicits user input to manually identify some true targets in the image. The verification is performed according to a reinforcement learning process to minimize a total verification time.

One variation of a method for stock keeping in a store includes: accessing an image captured by a fixed camera within the store; retrieving a field of view of the fixed camera; estimating a segment of an inventory structure in the store depicted in the image based on a projection of the field of view onto a planogram of the store; identifying a set of slots within the inventory structure segment; retrieving a product model representing a set of visual characteristics of a product type assigned to a slot, in the set of slots, by the planogram; extracting a constellation of features from the image; if the constellation of features approximates the set of visual characteristics in the product model, detecting presence of a product unit of the product type occupying the inventory structure segment; and representing presence of the product unit, occupying the inventory structure segment, in a realogram.

Non-limiting embodiments provide a registration system (10) including a detection unit (11) that detects an object in an image obtained by a camera imaging a placing surface of a table on which a product is placed, a recognition unit (12) that recognizes which product the object is, a first display unit (13) that displays first information for determining a first object which is an object not recognized by the recognition unit (12) on the placing surface, a second display unit (14) that displays product candidates for the first object on a display, a selection input reception unit (15) that receives a selection input for selecting one of the product candidates, and a registration unit (16) that registers the product recognized by the recognition unit (12) and the product candidate selected by the selection input as a checkout target.

In summary, to simplify operation of a medical image acquisition system, it is proposed that the image acquisition system continuously monitors a current image acquisition situation and assigns at least one new function to at least one, preferably manual, operating element, which is adapted to a detected new image acquisition situation, in response to a change in the image acquisition situation, to the extent that said newly assigned function is adjustable and/or operable with the at least one operating element. To this end, the image acquisition system can detect the change in the image acquisition situation using predefined parameters by means of sensors and/or through communication with peripheral units and/or preferably through image analysis of an image sequence which is recorded with an image sensor of the image acquisition system.

In a computer implemented sketch-based education or training system, guidance elements are generated and output to users both on an affirmative request of the user and in an automated manner without a request for guidance from the user. Automated guidance may take the form of a mini-hint that does not provide explicit information about a solution. The automatically provided guidance elements may contain numerical measures of correspondence between a user submitted sketch and a model sketch.

According to the embodiment, a product recognition apparatus includes a shooting device, a display, and a processor. The aforementioned processor controls the display so that a frame image of an object shot by the shooting device is displayed in an image display region of the display. Also, the processor controls the display so that a location image illustrating the location of the object is displayed so as to overlap the frame image in the image display region of the display.

An object-extraction method includes generating multiple partition objects based on an electronic document, and receiving a first user selection of a data element via a user interface of a compute device. In response to the first user selection, and using a machine learning model, a first subset of partition objects from the multiple partition objects is detected and displayed via the user interface. A user interaction, via the user interface, with one of the partition objects is detected, and in response, a weight of the machine learning model is modified, to produce a modified machine learning model. A second user selection of the data element is received via the user interface, and in response and using the modified machine learning model, a second subset of partition objects from the multiple partition objects is detected and displayed via the user interface, the second subset different from the first subset.

Systems and methods for analyzing, segmenting, and classifying multimedia material are disclosed herein. Embodiments include (i) receiving multimedia material for analysis, (ii) extracting elements from the multimedia material and forming objects comprising the elements; (iii) segmenting the multimedia material into segments, where individual segments include objects located within a threshold distance from each other; (iv) detecting objects within each segment; (v) associating attributes with the detected objects within the segments; (vi) annotating the segments by creating a relationship tree among the objects within each segment; and (vii) storing annotations of the segments for analysis.

An image processing method performed by a computing device deployed with a deep-learning neural network is provided. An image, including an object to be segmented from the image, is acquired. The object is segmented from the image by using the deep-learning neural network to acquire a first segmentation result. Correction information input by a user with respect to the first segmentation result is acquired. Based on correction information, the first segmentation result is modified by using the deep-learning neural network, to acquire a second segmentation result.