A device, method, and non-transitory computer readable medium are described. The method includes receiving a dataset including hand written Arabic words and hand written Arabic alphabets from one or more users. The method further includes removing whitespace around alphabets in the hand written Arabic words and the hand written Arabic alphabets in the dataset. The method further includes splitting the dataset into a training set, a validation set, and a test set. The method further includes classifying one or more user datasets from the training set, the validation set, and the test set. The method further includes identifying the target user from the one or more user datasets. The identification of the target user includes a verification accuracy of the hand written Arabic words being larger than a verification accuracy threshold value.

A device, method, and non-transitory computer readable medium are described. The method includes receiving a dataset including hand written Arabic words and hand written Arabic alphabets from one or more users. The method further includes removing whitespace around alphabets in the hand written Arabic words and the hand written Arabic alphabets in the dataset. The method further includes splitting the dataset into a training set, a validation set, and a test set. The method further includes classifying one or more user datasets from the training set, the validation set, and the test set. The method further includes identifying the target user from the one or more user datasets. The identification of the target user includes a verification accuracy of the hand written Arabic words being larger than a verification accuracy threshold value.

An integrated gaze tracker includes a pixel array configured to capture an image to provide an output signal comprising the image; a memory configured to store an image data unit to be read out in units of lines based on the output signal of the pixel array; and a processor configured to generate a gaze vector based on partial image data comprising the image data unit, among entire image data included in the output signal.

An information processing apparatus includes processing circuitry configured to register a set of item names corresponding to an item name field associated with an item value field in a predetermined business form, and identify an association of an item name field with an item value field included in a business form to be processed based on the set of item names registered.

Ideogram character analysis includes partitioning an original ideogram character into strokes and mapping each stroke to a corresponding stroke identifier (id) to create an original stroke id sequence that includes stroke identifiers. A candidate ideogram character that has a candidate stroke id sequence within a threshold distance to the original stroke id sequence is selected. One or more embodiments may create a new phrase by replacing the original ideogram character with the candidate ideogram character in a search phrase. One or more embodiments perform a search using the search phrase and the new phrase to obtain a result and present the result. One or more embodiments may replace an original ideogram character in a character recognized document with the candidate ideogram character and store the character recognized document.

A label processing engine receives, as inputs, raw data representative of a label and baseline data, the product label being unique to at least one specific product and the baseline data corresponding to the at least one specific product. The engine detects a raw data object within the raw data, classifies the raw data object, and localizes the raw data object within the raw data. The engine detects a baseline data object within the baseline data, classifies the baseline data object, and localizes the baseline data object within the baseline data. The engine recognizes corresponding text within the raw and baseline data objects and extracts the corresponding text, reassembles the corresponding text into respective lines of text, compares the respective lines of text with one another, and generates a tabular and/or visual report including a notification based on the comparison.

The present application provides a database generation method and apparatus, an electronic device and a medium. The method includes: acquiring an image of a Chinese character, determining, based on the image of the Chinese character, elementary components contained in the Chinese character and positions of the elementary components in the image of the Chinese character; determining position codes corresponding to the positions of the elementary components; and storing the elementary components and the corresponding position codes thereof in a corresponding manner to obtain a component library of a database. Thus, the difficulty of generating the database is reduced.

A Portable Document Format (PDF) document processing method, in the fields of natural language processing and computer vision in artificial intelligence, includes: performing Optical Character Recognition (OCR) on a PDF document in image form to obtain recognized target results and first coordinate information for each target result, where each target result includes one character or at least two consecutive characters forming a character segment; for each target result, converting the first coordinate information into corresponding PDF coordinates and using the PDF coordinates as second coordinate information; determining a target font library for rewriting based on one or more characters in each target result; rewriting the one or more characters from each target result into the PDF document based on the second coordinate information and the target font library, to obtain a desired target document.

A feature extraction system, method and apparatus based on neural network optimization by gradient filtering is provided. The feature extraction method includes: acquiring, by an information acquisition device, input information; constructing, by a feature extraction device, different feature extraction networks, performing iterative training on the networks in combination with corresponding training task queues to obtain optimized feature extraction networks for different input information, and calling a corresponding optimized feature extraction network to perform feature extraction according to a class of the input information; performing, by an online updating device, online updating of the networks; and outputting, by a feature output device, a feature of the input information. The new feature extraction system, method and apparatus avoids the problem of catastrophic forgetting of the artificial neural network in continuous tasks, and achieves high accuracy and precision in continuous feature extraction.

A first character string obtainment unit according to one or more embodiments may obtain a first character string in response to an input character string that has been input. A similar character extraction unit extracts similar characters having similar shapes as characters in the first character string. A second character string generation unit generates one or more second character strings in which some or all of the characters in the first character string are replaced with similar characters extracted by the similar character extraction unit. Then, a conversion candidate output unit outputs the first character string and the second character strings as conversion candidates for the input character string.