US 12,015,731 B2
Call classification through analysis of DTMF events
Nick Gaubitch, Atlanta, GA (US); Scott Strong, Atlanta, GA (US); John Cornwell, Atlanta, GA (US); Hassan Kingravi, Atlanta, GA (US); and David Dewey, Atlanta, GA (US)
Assigned to Pindrop Security, Inc., Atlanta, GA (US)
Filed by Pindrop Security, Inc., Atlanta, GA (US)
Filed on Jul. 5, 2022, as Appl. No. 17/857,618.
Application 17/857,618 is a continuation of application No. 17/157,848, filed on Jan. 25, 2021, granted, now 11,388,490.
Application 17/157,848 is a continuation of application No. 16/378,286, filed on Apr. 8, 2019, granted, now 10,904,643, issued on Jan. 26, 2021.
Application 16/378,286 is a continuation of application No. 15/600,625, filed on May 19, 2017, granted, now 10,257,591, issued on Apr. 9, 2019.
Claims priority of provisional application 62/370,135, filed on Aug. 2, 2016.
Claims priority of provisional application 62/370,122, filed on Aug. 2, 2016.
Prior Publication US 2022/0337924 A1, Oct. 20, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. H04M 1/56 (2006.01); H04L 25/02 (2006.01); H04M 3/22 (2006.01); H04M 3/493 (2006.01); H04M 7/12 (2006.01); H04M 15/06 (2006.01); H04Q 1/45 (2006.01); H04Q 3/70 (2006.01); G10L 25/51 (2013.01)
CPC H04M 1/56 (2013.01) [H04L 25/0202 (2013.01); H04M 3/2281 (2013.01); H04M 3/493 (2013.01); H04M 7/1295 (2013.01); H04Q 1/45 (2013.01); H04Q 3/70 (2013.01); G10L 25/51 (2013.01); H04M 2201/18 (2013.01); H04M 2203/60 (2013.01); H04Q 2213/13139 (2013.01); H04Q 2213/13405 (2013.01); H04Q 2213/13515 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
receiving, by a computer, dual-tone multifrequency (DTMF) information for a plurality of past calls;
for each past call, generating, by the computer, a feature vector based upon the DTMF information for the past call, the feature vector associated with a label corresponding to a characteristic of the past call;
training, by the computer, a classifier model according to each of the feature vectors and the associated label for each of the past calls; and
generating, by the computer, one or more fingerprint models associated with a known caller by executing the classifier model on one or more prior feature vectors generated using the DTMF information for one or more of past calls associated with the known caller.