Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Identifying a User of a Multi-User Speech Recognition System by Audio Recordings

IP.com Disclosure Number: IPCOM000123607D
Original Publication Date: 1999-Feb-01
Included in the Prior Art Database: 2005-Apr-05
Document File: 1 page(s) / 43K

Publishing Venue

IBM

Related People

Keller, A: AUTHOR [+2]

Abstract

Disclosed is a method for identifying the user of a speech recognition system having multiple users by allowing him to record his voice, which is subsequently played back during events in which it is critical to know the user for which the system is set.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 100% of the total text.

Identifying a User of a Multi-User Speech Recognition System by Audio
Recordings

   Disclosed is a method for identifying the user of a speech
recognition system having multiple users by allowing him to record
his voice, which is subsequently played back during events in which
it is critical to know the user for which the system is set.

   According to this method, each user of the system records
voice prompts, which are played back when he reaches a critical
point in the use of the speech recognition system, from which
further operation is dependent upon the system using his personal
speech file, or from which this speech file will be updated.  The
personal speech file of each user, which contains user-specific
information on his pronunciation of words, specialized vocabulary,
frequency of word usage, etc., is used to increase the accuracy of
speech recognition.

   The figure shows an exemplary display screen which is used
for recording voice prompts at the beginning of a session.  The list
of messages to record describe the critical events that will
subsequently provide the user with a subtle notification that he is
considered by the system to be its current user.  The user can
select the messages to record and can preview and change his
recordings.  When he subsequently hears a recorded prompt during a
session, he can readily determine whether the correct personal speech
file is being used by recognizing his own voice.