N10-1001 |
models to automatically segment
|
meeting recordings
|
. Meetings can be usefully segmented
|
H01-1051 |
downsample the wide-band signals of the
|
meeting recordings
|
. The language model contained
|
P07-1127 |
meeting ? 3 . Recordings : the
|
meeting recordings
|
( audio , visual , transcription
|
N10-1001 |
corpus contains over 100 hours of
|
meeting recordings
|
, with the different recording
|
N07-1004 |
four meetings . Each series of
|
meeting recordings
|
uses four distinctive speakers
|
N13-1109 |
consists of naturally occurring
|
meetings recordings
|
. This corpus is widely used
|
W04-2328 |
particularly relevant to a general-domain
|
meeting recording
|
application . The DA tags in
|
N10-1001 |
projects was a multimodal corpus of
|
meeting recordings
|
, annotated at a number of different
|
N06-1047 |
the challenges of summarizing
|
meeting recordings
|
: namely , that utterances can
|
N06-1047 |
summarization system applied to
|
meeting recordings
|
from the ICSI Meetings corpus
|
W04-2313 |
were also able to listen to the
|
meeting recordings
|
. 4.2 First Experiment : Annotation
|
N10-1006 |
high speech recognition errors on
|
meeting recordings
|
. In this paper we investigated
|
H01-1003 |
of a number of internal group
|
meeting recordings
|
( about one hour long each )
|
N06-2014 |
2004 ) , using a subset of ICSI
|
meeting recording
|
corpus ( Janin et al. , 2003
|
N03-2012 |
experiments are based on a subset of
|
meeting recordings
|
collected and transcribed by
|
W04-0710 |
references made to documents in
|
meeting recordings
|
by linking referring expressions
|
N10-1001 |
interpretation ( section 4 ) of multimodal
|
meeting recordings
|
, application prototypes ( section
|
H01-1051 |
be simultaneously recorded . A
|
meeting recording
|
infrastructure is also being
|
W04-2328 |
dialogue understanding for multiparty
|
meeting recording
|
and retrieval ap - plications
|
N12-1041 |
multi-party interaction found in
|
meeting recordings
|
, performing extractive summarization
|