Abstract: Techniques are provided for computing systems, methods and computer program products to produce efficient image-to-image translation by adapting unpaired datasets for supervised learning. A first model (a powerful model) may be defined and conditioned using unsupervised learning to produce a synthetic paired dataset from the unpaired dataset, translating images from a first domain to a second domain and images from the second domain to the first domain. The synthetic data generated is useful as ground truths in supervised learning. The first model may be conditioned to overfit the unpaired dataset to enhance the quality of the paired dataset (e.g. the synthetic data generated). A run-time model such as for a target device is trained using the synthetic paired dataset and supervised learning. The run-time model is small and fast to meet the processing resources of the target device (e.g. a personal user device such as a smart phone, tablet, etc.

Abstract: An image-processing device includes: a character string detection unit configured to detect a character string of a specific item in a first document image based on a feature amount of the displayed first document image among feature amounts which are recorded in advance based on a result of learning using a plurality of document images and indicate features of the character string of the item for each kind of document image and each specific item; and an output unit configured to output information regarding a correspondence relation indicating the same specific item between the first document image and a second document image displayed to correspond to the first document image.

Abstract: A reference data extraction unit extracts, from a photographic image from an imaging device that captures an image of an object to be recognized, an image of a reference image region serving as a reference and containing a detection subject in the object. A expanded data extraction unit extracts from the photographic image an image of an expanded-image region, which is an image region that includes the reference image region and is larger than the reference image region. A reduced data extraction unit extracts from the photographic image an image of a reduced-image region, which is an image region that includes the detection subject and is smaller than the reference image region, with the result that a portion of the object is outside of the region. A learning unit uses the extracted images of the image region to learn a dictionary.

Abstract: A computer-implemented method for implementing face recognition includes receiving training data including a plurality of augmented images each corresponding to a respective one of a plurality of input images augmented by one of a plurality of variations, splitting a feature embedding generated from the training data into a plurality of sub-embeddings each associated with one of the plurality of variations, associating each of the plurality of sub-embeddings with respective ones of a plurality of confidence values, and applying a plurality of losses including a confidence-aware identification loss and a variation-decorrelation loss to the plurality of sub-embeddings and the plurality of confidence values to improve face recognition performance by learning the plurality of sub-embeddings.

Abstract: A method for updating a parameter of a model, a distributed training system, and an electric device are related to a field of deep learning technologies. The method includes: obtaining a batch training period of batch training data to be trained for a model; increasing priorities of tasks ranked at a bottom in a sequence of gradient communication tasks for parameters of the model when the batch training period is greater than or equal to a preset period threshold; and performing a communication of gradients of the parameters and updating the parameters based on priorities of the gradient communication tasks for the parameters in the model.

Abstract: Some embodiments provide a non-transitory machine-readable medium that stores a program executable by a device. The program receives a request to process a file. The file includes a set of images of text. The program further converts the text in each image in the set of images into a set of machine-readable text. The program also uses a machine learning model to predict, based on the set of machine-readable text, whether the set of images of the file are images of pages that belong to a single document or images of pages that belong to different documents.

Abstract: Techniques facilitating causal reasoning for explanation of model predictions are provided. A system can generate one or more explanations of a machine learning model prediction. The one or more explanations can be based on causal relationships determined between feature data of a set of feature data and based on dataset point samples around a trace associated with the causal relationships.

Abstract: A first communication device may execute an output control process of externally outputting output information, the output information obtained using a public key of the first communication device, receive an authentication request from a second communication device that has obtained the public key, send a first authentication response to the second communication device via a wireless interface in a case where it is determined that a wireless connection has been established between the first communication device and a first access point, the first authentication response including first role information indicating that the first communication device is to serve a first role which is a role of sending first wireless setting information to the second communication device, and send the first wireless setting information to the second communication device.

Abstract: A metrology system includes one or more through-focus imaging metrology sub-systems communicatively coupled to a controller having one or more processors configured to receive a plurality of training images captured at one or more focal positions. The one or more processors may generate a machine learning classifier based on the plurality of training images. The one or more processors may receive one or more target feature selections for one or more target overlay measurements corresponding to one or more target features. The one or more processors may determine one or more target focal positions based on the one or more target feature selections using the machine learning classifier. The one or more processors may receive one or more target images captured at the one or more target focal positions, the target images including the one or more target features of the target specimen, and determine overlay based thereon.

Abstract: A printing apparatus determines, upon receiving information from a server that printing of a print job acquired from the server should be cancelled after the start of print processing of the print job, whether the cancellation should be accepted. If it is determined the cancellation should be accepted, execution of the print processing is cancelled and terminated and the server is notified, and if it is determined the cancellation should not be accepted, the server is notified of an execution result indicating that output of the print job has been completed. The printing apparatus further notifies the server of the execution result indicating that output of the print job has been completed when the printing of the print job has been successfully completed before cancellation and termination of the execution of the print processing, even in a case where it is determined that the cancellation should be accepted.

Abstract: A printing apparatus includes a reception unit configured to receive a print job, and an execution unit configured to execute printing of the print job received by the reception unit. The execution unit has a function of, in a case where a setting time or more elapses with a particular cause preventing the printing of the print job not being removed, canceling the printing of the print job. The execution unit has a function of canceling printing of a new print job received by the reception unit in a state where the setting time or more elapses with the particular cause not being removed.

Abstract: Provided is a system and method for matching documents by line item. In one example, the method may include receiving a target document from a target system and a source document from a source system which are linked, identifying a set of lines in the target document and a set of lines in the source document which cannot be matched based on initial conditions, determining, via a machine learning model, a target line from the identified set of lines in the target document matches a source line from the identified set of lines in the source document based on descriptions included in the target line and the source line, and updating the source system to indicate that the target line from the target document and the source line from the source document are a match.

Abstract: An image forming apparatus communicates with another device and includes a processor that controls the communication. A control unit controls the apparatus to operate in one of at least a first power saving state, a second power saving state, and a normal state. In the first power saving state, the apparatus is controlled so that an image forming unit, an image reading unit, and the control unit are not activated, and a communication device is activated. In the second power saving state, the apparatus is controlled so that at least one of the image forming unit and the image reading unit are not activated, and the communication device and the control unit are activated. In the normal state, the apparatus is controlled so that at least the image forming unit, the image reading unit, the communication device, and the control unit are activated.

Abstract: An electronic apparatus is provided. The electronic apparatus includes a user interface, a camera, a memory configured to store a first artificial intelligence model trained to obtain information on an emotion based on an input image, and a processor, connected to the user interface, the camera, and the memory, configured to control the electronic apparatus. The processor is configured to, based on text being input through the user interface, identify whether the text includes first information on a user's emotion, based on the text including the first information, obtain second information for the emotion by inputting the image obtained through the camera to the first artificial intelligence model, and identify a type of the user's emotion based on the first information obtained from the text and the second information obtained from the image.

Abstract: Identification means 71 identifies an object indicated by data by applying the data to a model learned by machine learning. Determination means 72 determines whether or not the data is transmission target data to be transmitted to a predetermined computer based on a result obtained by applying the data to the model. Data transmission means 73 transmits the data determined to be the transmission target data to the predetermined computer at a predetermined timing.

Abstract: A printing apparatus capable of conveying sheets to a sheet processing apparatus and to control the type of a sheet to be used as a cover sheet in a case where a book bound product is generated by switching back sheets each having an image printed, inverting the sheets and folding the inverted sheets, and to notify an error to a user based on the type of a sheet to be used as the cover sheet in a case where printing an image on the cover sheet is designated.

Abstract: An estimation apparatus includes an acquisition unit configured to acquire data about a domesticated animal identified by identification information that is transmitted by a client terminal, and an estimation unit configured to perform estimation by inputting the acquired data about a domesticated animal to a trained model generated by performing machine learning based on captured image data of domesticated animals and collected data about domesticated animals and provide, to the client terminal, an estimation result indicating a result of the estimation, and the client terminal includes a presenting unit configured to transmit a request for estimation together with identification information for identifying a domesticated animal targeted for estimation, receive the estimation result, and present body weight data about the domesticated animal targeted for estimation to a user.

Abstract: A method is disclosed. For example, the method executed by a processor of a multi-function device (MFD) includes executing a defect learning routine to identify defects, cataloging the defects based on a job function, a type of paper, and a machine state, receiving a job request, determining a known defect that has been catalogued based on the job function, the type of paper, and the machine state, and presenting a visualization of the known defect on a display of a user interface before executing the job request.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for performing intelligent extraction of information from a document. A computing module receives input data representing an image of a document. The module also receives context data for the document. The context data includes parameters that are descriptive of the document in the image. The module processes the input data and the context data to determine a complexity value that characterizes a level of complexity in identifying information to be extracted from the document. The system selects a machine-learning model to use in extracting information from the document. The model is selected based on the complexity value and from multiple candidate models. The system extracts information from the document using the selected model, including converting a portion of the image of the document that shows typed or handwritten text into a digitized text string.

Abstract: Systems and methods are configured to generate a set of related document objects for a predictive entity and/or to generate an optimal document sequence for a set of related document objects. In one embodiment, a set of related document objects for a predictive entity is generated by processing entity metadata features associated with the predictive entity using an entity-document correlation machine learning model, and an optimal document sequence is generated for the set of related document objects by processing the set of related document objects using a document sequence optimization machine learning model.