Systems and methods of sound-based fraud protection

Patent metadata

Field	Value
Publication number	US-10726850-B2
Application number	US-201916460618-A
Country	US
Kind code	B2
Filing date	Jul 2, 2019
Priority date	Mar 20, 2018
Publication date	Jul 28, 2020
Grant date	Jul 28, 2020

Abstract

The present disclosure describes systems and methods of fraud protection. A listening device may monitor ambient sound and detect one or more triggering sound patterns. The listening device may record the one or more triggering sound patterns and record a person's voice, and send the recordings to a server, wherein the server is configured to calculate a confidence level of fraud based on the one or more triggering sound patterns and the recording of the person's voice using a speech pattern model of the person, and to output an alert if the confidence level of fraud is greater than a threshold value.

First claim

What is claimed is:

1. A computer-implemented method of fraud protection for a computer-accessible account, comprising:
monitoring, by a listening device, ambient sound;
detecting, by the listening device, one or more triggering sound patterns from the ambient sound, the one or more triggering sound patterns including sound indicative of an initiation of an interaction between an initiator and a recipient;
in response to detecting the one or more triggering sound patterns, transmitting, by the listening device, to a server in communication with the listening device, the one or more triggering sound patterns and a recording of the recipient's voice obtained by the listening device after the detecting,
and wherein the server is configured to:
determine a correlation between the one or more triggering sound patterns and at least one portion of speech from the recorded recipient's voice;
detect a difference in a speech pattern model of the recipient from the recorded recipient's voice, including determining that a speech pattern of the recipient is different from the speech pattern model of the recipient, the speech pattern model being based on one or more speech samples of the recipient;
calculate a confidence level of fraud being attempted against the recipient based on the determined correlation and the detected difference; and
generate and transmit an alert in response to determining that the confidence level of fraud is greater than a threshold value.

2. The method of claim 1 , wherein the one or more triggering sound patterns comprise at least one of “IRS,” “owe,” “money,” “debt,” “wire,” “IRA,” “investment,” “may be at risk,” “retirement,” “back taxes,” “act now,” “before it's too late,” “time is running out,” “failure to act,” “once in a lifetime,” “you've won,” “last chance,” “arrest,” or “pay.”

3. The method of claim 1 , wherein the one or more triggering sound patterns comprises at least one of a doorbell sound, a door knock sound, a phone ring, a voicemail, or a TV sound clip.

4. The method of claim 1 , wherein a duration of the recording is between 0.1 and 300 seconds.

5. The method of claim 1 , further comprising:
receiving, by the listening device, the alert from the server; and
outputting a notification in response to receiving the alert.

6. The method of claim 1 , further comprising:
detecting one or more triggering words in the recipient's voice; and
adjusting a duration of the recording in response to detecting the one or more triggering words.

7. A computer-implemented method of fraud protection for a computer-accessible account, comprising:
receiving, by a server, one or more speech samples associated with a person;
generating a speech pattern model of the person based on extracting a person's speech from the one or more speech samples;
storing the speech pattern model of the person;
receiving one or more triggering sound patterns and a recording of the person's voice obtained by a listening device, the one or more triggering sound patterns including sound indicative of an initiation of an interaction between an initiator and a recipient;
determining, a correlation between the one or more triggering sound patterns and at least one portion of speech from the person's voice, thereby identifying the person as the recipient;
detecting a difference in the speech pattern model from the recorded recipient's voice, including determining that a speech pattern of the recipient is different from the speech pattern model of the recipient;
determining a confidence level of fraud being attempted against the recipient from the determined correlation between the one or more triggering sound patterns and the recording of the recipient's voice and the detected difference; and
generating and transmitting an alert in response to determining that the confidence level of fraud is greater than a threshold value.

8. The method of claim 7 , wherein the one or more triggering sound patterns comprise at least one of “IRS,” “owe,” “money,” “debt,” “wire,” “IRA,” “investment,” “may be at risk,” “retirement,” “back taxes,” “act now,” “before it's too late,” “time is running out,” “failure to act,” “once in a lifetime,” “you've won,” “last chance,” “arrest,” or “pay.”

9. The method of claim 7 , wherein the one or more triggering sound patterns comprises at least one of a doorbell sound, a door knock sound, a phone ring, a voicemail, or a TV sound clip.

10. The method of claim 7 , wherein transmitting the alert comprises sending an alert to a pre-determined entity associated with the recipient.

11. The method of claim 10 , wherein the pre-determined entity comprises at least one of a financial institution, a law enforcement agency, a family member of the recipient, or a caregiver associated with the recipient.

12. The method of claim 7 , further comprising:
updating a flag status of the computer-accessible account associated with the recipient in response to determining that the confidence level of fraud exceeds a second threshold value; and
outputting the flag status of the account in a user interface.

13. The method of claim 12 , wherein the computer-accessible account associated with the recipient comprises at least one of a bank account, a checking account, a saving account, a credit card account, an investment account, a loan account, or an individual retirement account.

14. A system of fraud protection for a computer-accessible account, comprising:
one or more microphones configured to monitor ambient sound and detect one or more triggering sound patterns from the ambient sound, the one or more triggering sound patterns including sound indicative of an initiation of an interaction between an initiator and a recipient; and
a processor circuit coupled to the one or more microphones and configured to execute instructions causing the processor to:
in response to the one or more microphones detecting the one or more triggering sound patterns, transmit, to a server in communication with the listening device, the one or more triggering sound patterns and a recording of the recipient's voice obtained by the one or more microphones after the detecting;

and wherein the server is configured to:
determine a correlation between the one or more triggering sound patterns and the recording of the recipient's voice;
determine a speech pattern of the person from the recorded recipient's voice;
determine a correlation between the one or more triggering sound patterns and at least one portion of speech from the recorded recipient's voice;
detect a difference in a speech pattern model of the recipient from the recorded recipient's voice, including determining that the speech pattern of the recipient is different from the speech pattern model of the recipient, the speech pattern model being based on one or more speech samples of the recipient;
calculate a confidence level of fraud being attempted against the recipient based on the determined correlation and the detected difference; and
generate and transmit an alert in response to determining that the confidence level of fraud is greater than a threshold value.

15. The system of claim 14 , wherein the one or more triggering sound patterns comprise at least one of “IRS,” “owe,” “money,” “debt,” “wire,” “IRA,” “investment,” “may be at risk,” “retirement,” “back taxes,” “act now,” “before it's too late,” “time is running out,” “failure to act,” “once in a lifetime,” “you've won,” “last chance,” “arrest,” or “pay.”

16. The system of claim 14 , wherein a duration of the recording is between 0.1 and 300 seconds.

17. The sy

Assignees

• Capital One Services Llc

Inventors

• Mossoba Michael
• Edwards Joshua
• Cunningham Sarah
• Rothbaum Mark

Classifications

• H04W12/65

  Environment-dependent, e.g. using captured environmental data · CPC title

• G10L17/00

  Speaker identification or verification techniques · CPC title

• G10L25/48

  specially adapted for particular use · CPC title

• G10L25/51

  for comparison or discrimination · CPC title

• G10L15/26

  Speech to text systems (G10L15/08 takes precedence) · CPC title