US 11,868,709 B1
Digitally aware neural dictation interface
Gerard K. Cohen, Roseville, CA (US); Edmund Gauci, San Jose, CA (US); and Eric Miller, Burlingame, 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 Oct. 22, 2019, as Appl. No. 16/660,656.
Application 16/660,656 is a continuation of application No. 16/600,242, filed on Oct. 11, 2019.
Int. Cl. G06F 40/00 (2020.01); G06F 40/174 (2020.01); G06F 3/16 (2006.01); G10L 17/02 (2013.01); G10L 17/06 (2013.01); G10L 17/00 (2013.01)
CPC G06F 40/174 (2020.01) [G06F 3/167 (2013.01); G10L 17/00 (2013.01); G10L 17/02 (2013.01); G10L 17/06 (2013.01)] 16 Claims
OG exemplary drawing
 
1. A method, comprising:
determining, by a processing circuit, characteristics of a plurality of fields of an electronic form based on executing a metadata analysis function, wherein the metadata analysis function comprises identifying the plurality of fields and identifying a position of each of the plurality of fields in relation to each other;
determining, by the processing circuit, a priority order of navigating to and populating the plurality of fields of the electronic form;
receiving, by the processing circuit, a first speech input from a user to enable a hands-free mode of operation;
authenticating, by the processing circuit, the user by comparing the first speech input from the user with known voice samples of the user;
in response to authenticating the first speech input based on the comparison, enabling, by the processing circuit, the hands-free mode of operation;
receiving, by the processing circuit in the hands-free mode of operation, a second speech input for filling out a first field of the plurality of fields of the electronic form, wherein the first field is selected based on the priority order;
determining, by the processing circuit, an irregularity in the second speech input based on a characteristic of the first field and identifying a missing syllable and a distorted syllable in the second speech input, wherein identifying the missing syllable and the distorted syllable comprises executing a syllable-by-syllable analysis of the second speech input, wherein the missing syllable is determined based on other syllables identified in the syllable-by-syllable analysis, and wherein the distorted syllable is determined based on failing to recognize a syllable in the syllable-by-syllable analysis;
refining, by the processing circuit, the second speech input into at least one matching syllable by extrapolating the missing syllable or the distorted syllable based on stored syllables of a plurality of past speech inputs received from the user, wherein the at least one matching syllable is selected based on comparing each syllable in the second speech input to syllables stored in a database to find a match for the missing syllable or the distorted syllable in the second speech input, and wherein the match is determined at least in part on an expected field value associated with the first field;
converting, by the processing circuit, the refined second speech input comprising the matching syllable into text; and
providing, by the processing circuit, the text to a user device to populate the first field with the text.