US 11,854,286 B2
Image-based document analysis using neural networks
Steven Gilbert, New York, NY (US); and Mugundhan Elamathi, New York, NY (US)
Assigned to JPMORGAN CHASE BANK , N.A., New York, NY (US)
Filed by JPMORGAN CHASE BANK, N.A., New York, NY (US)
Filed on Sep. 20, 2022, as Appl. No. 17/933,795.
Application 17/933,795 is a continuation of application No. 16/867,441, filed on May 5, 2020, granted, now 11,568,663.
Prior Publication US 2023/0011841 A1, Jan. 12, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06V 30/413 (2022.01); G06F 16/93 (2019.01); G06F 16/54 (2019.01); G06F 16/11 (2019.01); G06N 3/08 (2023.01); G06Q 10/10 (2023.01)
CPC G06V 30/413 (2022.01) [G06F 16/116 (2019.01); G06F 16/54 (2019.01); G06F 16/93 (2019.01); G06N 3/08 (2013.01); G06Q 10/10 (2013.01)] 21 Claims
OG exemplary drawing
 
1. A computer-implemented method for scoring an entity, the method comprising the steps of:
accessing a set of documents, each document being associated with an entity identifier, each document being formatted in a text file format;
generating a set of graphical images by converting each document to a corresponding graphical image having a graphical file format;
inputting the set of graphical images into a neural network; receiving, as a first output of the neural network, a score that is assigned to the set of documents; and receiving, as a second output of the neural network, an output image, the output image being generated by modifying a visual characteristic of a region of at least one of the graphical images among the set of graphical images, wherein the region comprises a hotspot indicating an outlier of a metric represented in the set of documents.