HP3000-L Archives

October 1997, Week 4

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Nick Demos <[log in to unmask]>
Reply To:
Nick Demos <[log in to unmask]>
Date:
Wed, 22 Oct 1997 11:25:20 -0400
Content-Type:
text/plain
Parts/Attachments:
text/plain (31 lines)
I have been watching with interest all these comments about
voice recognition, "natural" languages and translations.  A few
comments:

1.  Voice recognition is not a translation, but just what it says -
    Recognizing words.

2.  of course there are hononyms (words that sound the same).  There
    come to mind two techniques that ameliorate this problem:
    A.  A selection process which appears on the screen for
    an unregnized or ambiguous word - the user either selects an
    alternative or types in a word.
    B.  The computer remembers this and presents this word at the
    top of the selection the next time the same sound is heard.

3.  There would have to some escape word/sound so that the computer
    would treat the following word as a command rather than as a
    text word.  This might also be accomplshed by holding a key
    down.

In summary voice recognition at the start could be very useful even
if not perfect and once widely used this use would fund improvements.

The same coould be said, BTW, for computer language translation.

Nick Demos  [log in to unmask]
My opinions are ny own and I stand behind them.

Performance Software Group
Tel. (410) 788-6777 Fax (410) 788-4476

ATOM RSS1 RSS2