Browse Prior Art Database

Quick correction of dictation result by recording user's action with effective replay

IP.com Disclosure Number: IPCOM000123141D
Original Publication Date: 1998-Jun-01
Included in the Prior Art Database: 2005-Apr-04
Document File: 2 page(s) / 58K

Publishing Venue

IBM

Related People

Iuchi, N: AUTHOR [+3]

Abstract

A program is disclosed that correction method of dictated text with listening user's voice. In current way, an user have to detect the mis-recognized word by reading the dictated text from the top of document, after recognition is completed. It is sometimes inconvenient because user often forgets what he/she says, especially the dictated document is long.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 73% of the total text.

Quick correction of dictation result by recording user's action with
effective replay

   A program is disclosed that correction method of dictated
text with listening user's voice.  In current way, an user have to
detect the mis-recognized word by reading the dictated text from the
top of document, after recognition is completed.  It is sometimes
inconvenient because user often forgets what he/she says, especially
the dictated document is long.

   In the subject method, user can detect the mis-recognized
word with playback of spoken voice and the text is displaying
simultaneously like a news ticker.  This calls "Correction marking
window" (see figure).

   The operation is following:  1) Select "Correction Start"
menu on the dictation word-processor, then Correction marking window
appears.  2) Playback user's voice and display the text that is
scrolling simultaneously.  3) If user detect the mis-recognized word,
push space key.  4) Change the display color of the word selected by
user (marking the word), and continue playback.  5) All the text
playback is finished, start correction automatically for all of the
word which are marked by user in operation 3).

   The actual correction operation is required after marking
operation is completed.  So user can concentrate if the word is
correctly dictated or not during playback the text.

   In the operation 2), to synchronize the playback and
display the scrolling speed is changed on each word, the speed is
calcula...