<s>
TIMIT	B-Application
is	O
a	O
corpus	O
of	O
phonemically	O
and	O
lexically	O
transcribed	O
speech	O
of	O
American	O
English	O
speakers	O
of	O
different	O
sexes	O
and	O
dialects	O
.	O
</s>
<s>
TIMIT	B-Application
was	O
designed	O
to	O
further	O
acoustic-phonetic	O
knowledge	O
and	O
automatic	B-Application
speech	I-Application
recognition	I-Application
systems	O
.	O
</s>
<s>
There	O
is	O
also	O
a	O
telephone	O
bandwidth	O
version	O
called	O
NTIMIT	B-Application
(	O
Network	O
TIMIT	B-Application
)	O
.	O
</s>
<s>
TIMIT	B-Application
and	O
NTIMIT	B-Application
are	O
not	O
freely	O
available	O
—	O
either	O
membership	O
of	O
the	O
Linguistic	O
Data	O
Consortium	O
,	O
or	O
a	O
monetary	O
payment	O
,	O
is	O
required	O
for	O
access	O
to	O
the	O
dataset	O
.	O
</s>
<s>
The	O
TIMIT	B-Application
telephone	O
corpus	O
was	O
an	O
early	O
attempt	O
to	O
create	O
a	O
database	O
with	O
speech	O
samples	O
.	O
</s>
<s>
It	O
was	O
published	O
in	O
the	O
year	O
1988	O
on	O
CD-ROM	B-Device
and	O
consists	O
of	O
only	O
10	O
sentences	O
per	O
speaker	O
.	O
</s>
<s>
The	O
full	O
name	O
of	O
the	O
project	O
is	O
DARPA-TIMIT	O
Acoustic-Phonetic	O
Continuous	O
Speech	O
Corpus	O
and	O
the	O
acronym	O
TIMIT	B-Application
stands	O
for	O
Texas	O
Instruments/Massachusetts	O
Institute	O
of	O
Technology	O
.	O
</s>
<s>
The	O
main	O
reason	O
why	O
a	O
corpus	O
of	O
telephone	O
speech	O
was	O
created	O
was	O
to	O
train	O
speech	B-Application
recognition	I-Application
software	I-Application
.	O
</s>
<s>
In	O
the	O
Blizzard	O
challenge	O
,	O
different	O
software	O
has	O
the	O
obligation	O
to	O
convert	O
audio	O
recordings	O
into	O
textual	O
data	O
and	O
the	O
TIMIT	B-Application
corpus	O
was	O
used	O
as	O
a	O
standardized	O
baseline	O
.	O
</s>
