Prekladatelske dotazy

[Ladislav Dobias]

> >  This is a very small diphone database, which gives poor quality soun=
d
> >                       ^^^^^^^
> >  output but runs very fast.
>=20
>    Datab=E1ze dvojzvuk=F9.
              ^^^^^^^^^
To ne - to je nesrozumitelne.

Ja psal clanek o rozpoznavani reci a tam to bylo
neprelozene (difo'ny - s dlouhym o). A myslim, ze puvodni=20
autor byl dosti verohodny (tusim pan Psutka - kniha Umela
inteligence (1/2?), popr. Metody rozpoznavani reci nebo tak
nejak).

Pro informaci:
Difony jsou segmenty reci odpovidajici useku ze stredu
samohlasky do stredu nasledne souhlasky a naopak.

Zdravi
Lada Dobias

--

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
                                                      ,v
  ##       ##   #####   #####            Ladislav DOBIAS
  ##      ####  ##  ## ##O-O##         Xdobiasl@fel.cvut.cz
  ##     ##  ## ##  ## ## > ##   <http://cs.felk.cvut.cz/~xdobiasl/>
  ##     ###### ##  ## ## v ##   I study Czech Technical University,
  ###### ##  ## #####   #####      Faculty of Electrical Engeneering

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%