Automation Anywhere, Inc.

Derek S Chan

Abhijit Kakhandiki

Nishit Kumar

Thomas Corcoran

Nakuldev Patel

Bruno Selva

Yogesh Savalia

Mukesh Methaniya

Priti Sundar Hazra

Information contained in tables in a digitized document is extracted by retrieving table layout data regarding bounding boxes, each being auto-generated by the system and/or (re)generated by a user to the digitized image of a sample document. A row template is used to identify a first table, by automatically scanning within the document. Upon detecting a possible row in the input image, a Row Possibility Confidence Value (RPCV) is generated that indicates a likelihood that the possible row corresponds to an actual row in the first table. The possible row is regarded as an actual row if the RPCV exceeds a predetermined threshold value. For repeated tables in a document only the first table needs to be identified via bounding boxes. Also, related tables can be linked to permit linked data to be extracted to a structured file. Also, only the primary column in a readable and existent table header is required to extract table values across columns.

Computerized recognition and extraction of tables in digitized documents

A system that automatically recognizes checkboxes within a document and extracts the contents thereof. A user views a digitized document and annotates the document by identifying checkboxes contained in the document, by way of visually perceptible bounding boxes. The annotated document is processed by a machine learning engine that employs multiple convolutional operations followed by a global average pooling layer, a fully connected layer with 1024 node and 'ReLU' activation, a fully connected layer with 2 node and 'softmax' activation. The identified checkboxes and their contents are stored as label-value pairs, where the label identifies the checkbox and the value identifies the value of the checkbox, which can be either Yes, No, or No checkbox found.

Computerized recognition of checkboxes in digitized documents

Key/Value pairs, each comprising a keyword string and an associated value, are extracted automatically from a document image. Each document image has a plurality of pixels with each pixel having a plurality of bits. A first subset of the plurality of bits for each pixel represents information corresponding to the document image. The document image is processed to add information to a second subset of the plurality of bits for each pixel. The information added to the second subset alters the appearance of the document image in a manner that facilitates semantic recognition of textually encoded segments within the document image by a Deep Neural Network (DNN) trained to recognize images within image documents. The DNN detects groupings of text segments within detected spatial templates within the document image. The text segments are mapped to known string values to generate the keyword strings and associated values.

Automatic key/value pair extraction from document images using deep learning

An optical character recognition system employs a deep learning system that is trained to process a plurality of images within a particular domain to identify images representing text within each image and to convert the images representing text to textually encoded data. The deep learning system is trained with training data generated from a corpus of real-life text segments that are generated by a plurality of OCR modules. Each of the OCR modules produces a real-life image/text tuple, and at least some of the OCR modules produce a confidence value corresponding to each real-life image/text tuple. Each OCR module is characterized by a conversion accuracy substantially below a desired accuracy for an identified domain. Synthetically generated text segments are produced by programmatically converting text strings to a corresponding image where each text string and corresponding image form a synthetic image/text tuple.

Growing community of inventors

San Jose, CA, United States of America

Derek S Chan