US 12,223,295 B2
Augmented reality powered auto code generator
Arunkumar Somasundaram, Coimbatore (IN); Udhayan M, Madurai (IN); Lalitha Rajeswari Karanam, Hyderabad (IN); and Balamurali Lakshminarayanan, Chennai (IN)
Assigned to Bank of America Corporation, Charlotte, NC (US)
Filed by Bank of America Corporation, Charlotte, NC (US)
Filed on Apr. 25, 2023, as Appl. No. 18/138,779.
Prior Publication US 2024/0361995 A1, Oct. 31, 2024
Int. Cl. G06F 8/30 (2018.01); G06F 8/36 (2018.01); G06F 8/71 (2018.01); G06T 19/00 (2011.01)
CPC G06F 8/36 (2013.01) [G06F 8/71 (2013.01); G06T 19/006 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An augmented reality (“AR”) device for auto-generating executable programming code, the AR device being part of a network of AR devices, the AR device comprising:
a camera operable to continuously capture an image of data displayed within pre-determined boundaries of a display;
a microprocessor operable to:
receive each captured image from the camera and store each captured image at the AR device;
analyze each captured image in order to identify a captured image comprising a plurality of text associated with computer processing steps for processing a set of documents;
in response to an identification of the captured image, transmit an electronic communication to the network of AR devices for retrieval of a most-repeated image captured by each of a remaining AR devices within the network of AR devices; and
simultaneous to a transmittal of the electronic communication, identify a most-repeated image from the AR device by:
retrieving each captured image stored at the AR device that has been historically received within a pre-determined time period prior to a receipt of the captured image;
analyzing each captured image received prior to the captured image for a most-repeated image;
in response to the analyzing, identifying the most-repeated image, wherein the most-repeated image comprises the set of documents for processing;
comparing the most-repeated image from the AR device to the most-repeated image retrieved from each of the remaining AR devices; and
when at least one of a most repeated images retrieved comprises the set of documents, determining a classifier for the set of documents, the classifier defining a document-type for the set of documents;
an artificial intelligence (“AI”) image processing neural network application configured to convert the captured image into a text file;
an AI analyzer application configured to identify within the text file a set of computer processing steps; and
an AI code generator application configured to generate code to perform the computer processing steps, the AI code generator configured to:
search a production repository for a pre-generated function that performs the set of computer processing steps based on the classifier;
when the pre-generated function in the production repository is identified, transmit the pre-generated function to an AI code builder for compiling; and
when the pre-generated function in the production repository is not identified:
search a plurality of code containers to find an optimal function for each step in the set of computer processing steps; and
in response to a selection of the optimal function for each step, transmit the optimal function for each step to an AI code builder for compiling.