Autor: Piotr Kozbial (piotrk_at_ibbrain.ibb.waw.pl)
Data: Sat 30 Mar 1996 - 04:09:52 MET
> Ale w przyszlosci - kto wie? - za pare lat moze bedzie mozna dyktowac
> komputerowi po polsku (na razie tylko po angielsku)..
>
> Witek P.
Nie za pare lat lecz na jesieni ma pojawic sie program (+ uzadzenie
wielkosci wspolczesnych napedow CD-ROM) do rozpoznawania glosu, ktory
jest duzo lepszy od dzis dostepnych. Jest on nie tylko lepszy (rozpoznaje
glos kilku ludzi mowiacych jednoczesnie), ale bedzie tez wymagal duzo
slabszych komputerow od dotychczas wymaganych do "Voice recognition".
Ponizej przesylam krotki fragment na ten temat. A jesli ktos bardzo chce
to moge mu przeslac caly tekst (jesli bedzie wiecej niz 3 chetnych to
wysylam to na liste).
Piotr
--------------------------------------------------------------------
SALT LAKE CITY, March 27 /PRNewswire/
-- For years viewers have been awed by the ability of their favorite
sci-fi characters to converse with their computers.Now, thanks to a
dramatic breakthrough, fonix(TM) corporation is bridging the gap between
man and machine through the use of new proprietary, patent-pending speech
recognition technologies. The new technologies, which should be
available for end-users by fourth quarter 1996, are in final
development stages before Beta testing begins.
(...)
A Genuine Breakthrough - fonix technologies are unlike any other voice
recognition technology currently on the market. These breakthrough
technologies allow for multiple users with no enrollment whatsoever,
unrestricted speech, and normal speaking speeds and phrasing at a minimum
of 120 words per minute using minimal host computer capacity.
--------------------------------------------------------------------
To archiwum zostało wygenerowane przez hypermail 2.1.7 : Tue 18 May 2004 - 12:41:21 MET DST