# US 2025/0173462 A1 · Securing User-Entered Text In-Transit

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2025/0173462 A1
- Document ID: US20250173462A1
- Status: Published application
- Kind: A1
- Application number: 18/971,362
- Filed: 2024-12-06
- Published: 2025-05-29
- Earliest claimed: 2019-11-15
- Assignees: Capital One Services, LLC
- Applicants: Capital One Services, LLC
- Inventors: Jeremy Goodsitt; Austin Walters; Galen Rafferty
- Web page: https://patentgrep.com/patent/US20250173462A1

## CPC

- G06F 21/6263: {during internet communication, e.g. revealing personal data from cookies}
- G06F 18/2413: based on distances to training or reference patterns
- G06F 21/602: {Providing cryptographic facilities or services}
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/047: Probabilistic or stochastic networks
- G06N 3/0475: Generative networks
- G06N 3/09: Supervised learning
- G06N 3/094: Adversarial learning
- G06N 3/096: Transfer learning
- G06V 10/764: using classification, e.g. of video objects
- G06V 10/82: using neural networks
- G06N 3/045: Combinations of networks
- G06N 3/082: modifying the architecture, e.g. adding, deleting or silencing nodes or connections
- G06N 3/084: Backpropagation, e.g. using gradient descent
- G06N 3/088: Non-supervised learning, e.g. competitive learning
- G06V 30/413: Classification of content, e.g. text, photographs or tables
- G06V 40/376: {Acquisition}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

Systems and methods described herein discuss securing user-entered data in-transit between a first device and a second device. A user may enter text in a document. A first device may analyze the document to identify the user-entered text. The user-entered text may be separated from the document and transformed into an image using a machine learning algorithm. Transforming the text into an image may secure the data in-transit from the first device to a second device. The second device may receive the image and the document from the first device. The second device may reconstruct the user-entered text from the received image and re-assemble the document from the received document and the reconstructed user-entered text.

## Claims

### Claim 1

A computer-implemented method comprising:
identifying, based on an automated analysis of a document, a plurality of fields of user-entered text, wherein at least one field, of the plurality of fields, comprises personally identifiable information;
removing the plurality of fields from the document;
transforming, using a neural network, each field, of the plurality of fields, into an image that obfuscates the respective field while the document is in-transit to a second computing device;
generating a plurality of tags, wherein each tag, of the plurality of tags, indicates a respective location of each field; and
transmitting, to the second computing device, an electronic communication comprising the document with the plurality of fields removed, the plurality of images, and the plurality of tags.

### Claim 2 (depends on 1)

The computer-implemented method of claim 1, comprising:
encrypting, prior to the transmitting the electronic communication, the plurality of images; and
transmitting, to the second computing device, the plurality of encrypted images.

### Claim 3 (depends on 1)

The computer-implemented method of claim 1, further comprising:
identifying, during the automated analysis of the document, the plurality of fields of user-entered text using at least one of:
image segmentation; or.
background subtraction.

### Claim 4 (depends on 1)

The computer-implemented method of claim 1, further comprising:
transforming a first field of user-entered text into a first image using at least one of:
an image-to-image transformation;
a text-to-image transformation; or
a font transformation.

### Claim 5 (depends on 1)

The computer-implemented method of claim 1, wherein the neural network comprises at least one of:
a generative adversarial network (GAN);
a consistent adversarial network (CAN);
a cyclic generative adversarial network (C-GAN);
a deep convolutional GAN (DC-GAN);
GAN interpolation (GAN-INT);
GAN-CLS; or
a cyclic-CAN (C-CAN).

### Claim 6

A first computing device comprising:
one or more processors; and
memory storing instructions that, when executed by the one or more processors, cause the computing device to:
identify, based on an automated analysis of a document, a plurality of fields of user-entered text, wherein at least one field, of the plurality of fields, comprises personally identifiable information;
remove the plurality of fields from the document;
transform, using a neural network, each field, of the plurality of fields, into an image that obfuscates the respective field while the document is in-transit to a second computing device;
generate a plurality of tags, wherein each tag, of the plurality of tags, indicates a respective location of each field; and
transmit, to the second computing device, an electronic communication comprising the document with the plurality of fields removed, the plurality of images, and the plurality of tags.

### Claim 7 (depends on 6)

The first computing device of claim 6, wherein the instructions, when executed by the one or more processors, cause the first computing device to:
encrypt, prior to transmitting the electronic communication, the plurality of images; and
transmitting, to the second computing device, the plurality of encrypted images.

### Claim 8 (depends on 6)

The first computing device of claim 6, wherein the instructions, when executed by the one or more processors, cause the first computing device to:
identify, during the automated analysis of the document, the plurality of fields of user-entered text using at least one of:
image segmentation; or.
background subtraction.

### Claim 9 (depends on 6)

The first computing device of claim 6, wherein the instructions, when executed by the one or more processors, cause the first computing device to:
transform a first field of user-entered text into a first image using at least one of:
an image-to-image transformation;
a text-to-image transformation; or
a font transformation.

### Claim 10 (depends on 6)

The first computing device of claim 6, wherein the neural network comprises at least one of:
a generative adversarial network (GAN);
a consistent adversarial network (CAN);
a cyclic generative adversarial network (C-GAN);
a deep convolutional GAN (DC-GAN);
GAN interpolation (GAN-INT);
GAN-CLS; or
a cyclic-CAN (C-CAN).

### Claim 11

A computer-implemented method comprising:
receiving, by a second computing device and from a first computing device, a document with a plurality of fields of user-entered text removed, a plurality of images, and a plurality of tags wherein each tag, of the plurality of tags, indicates a location of a field of user-entered text in the document;
transforming, using a neural network, each image, of the plurality of images, into a respective field of user-entered text;
reconstructing, using the plurality of tags, a complete document by combining the respective fields of user-entered text and the received document;
comparing the plurality of fields of user-entered text to previously registered user data to determine whether the user-entered text matches the previously registered user data; and
verifying, based on a determination that the user-entered text matches the previously registered user data, the complete document.

### Claim 12 (depends on 11)

The computer-implemented method of claim 11, wherein the plurality of images are encrypted.

### Claim 13 (depends on 11)

The computer-implemented method of claim 11, comprising:
decrypting, prior to transforming each image into a respective field of user-entered text, the plurality of encrypted images.

### Claim 14 (depends on 11)

The computer-implemented method of claim 11, wherein transforming each image, of the plurality of images, into a respective field of user-entered text comprises at least one of:
an image-to-image transformation;
an image-to-text transformation; or
a font transformation.

### Claim 15 (depends on 11)

The computer-implemented method of claim 11, wherein the neural network comprises at least one of:
a generative adversarial network (GAN);
a consistent adversarial network (CAN);
a cyclic generative adversarial network (C-GAN);
a deep convolutional GAN (DC-GAN);
GAN interpolation (GAN-INT);
GAN-CLS; or
a cyclic-CAN (C-CAN).

### Claim at position 16

16-20. (canceled)
