US 11,935,524 B1
Authentication of impaired voices
Andrew J. Garner, IV, State Road, NC (US); Tyua Larsen Fraser, Livermore, CA (US); Kimberly Ann MacInnis, San Francisco, CA (US); Paul R. McMahon, Dayton, MN (US); Darrell Lee Suen, San Ramon, CA (US); and Zhong Wan, Pacifica, CA (US)
Assigned to Wells Fargo Bank, N.A., San Francisco, CA (US)
Filed by Wells Fargo Bank, N.A., San Francisco, CA (US)
Filed on Jan. 18, 2021, as Appl. No. 17/248,282.
Application 17/248,282 is a continuation of application No. 15/711,022, filed on Sep. 21, 2017, granted, now 10,896,673.
This patent is subject to a terminal disclaimer.
Int. Cl. G10L 15/20 (2006.01); G06F 21/32 (2013.01); G10L 17/08 (2013.01); G10L 17/20 (2013.01); G10L 17/22 (2013.01); G10L 17/24 (2013.01)
CPC G10L 15/20 (2013.01) [G06F 21/32 (2013.01); G10L 17/20 (2013.01); G10L 17/22 (2013.01); G10L 17/24 (2013.01); G10L 17/08 (2013.01)] 17 Claims
OG exemplary drawing
 
1. An authentication system for impaired voice authentication, the system comprising;
at least one processor; and
memory including instructions that, when executed by the at least one processor, cause the at least one processor to:
generate an initial voice profile for a user;
collect a plurality of voice samples of the user;
identify changes between the initial voice profile and the plurality of voice samples;
generate an alternate voice profile for the user using the initial voice profile and the identified changes;
transmit an inaudible audio tone pattern to a device of the user;
receive an audio stream including an authentication voice of the user; and
authenticate the authentication voice based on a match with a first set of authentication criteria and based on detection of the inaudible audio tone pattern in the audio stream, wherein the first set of authentication criteria includes acoustical property features of the initial voice profile and acoustical property features of the identified changes from the alternate voice profile.