prev1
stringlengths 1
7.9k
| prev2
stringlengths 1
7.9k
⌀ | next1
stringlengths 1
7.9k
| next2
stringlengths 1
7.9k
⌀ | is_boundary
bool 2
classes | session_id
stringlengths 7
14
|
|---|---|---|---|---|---|
I mean there 's a bunch of things in this note to various people
|
uh Broadcast News.
|
especially I guess um with Jane that that would help for since we have this new data now uh in order to go from the transcripts more easily to um just the words that the recognizer would use for scoring. I had to deal with some of it by hand but I think a lot of it can be automated s by
|
Oh one thing I guess I didn't get so you know the language model was straight from from bigram from Switchboard the acoustic models were also from Switchboard or or
| false |
QMSum_225
|
uh Broadcast News.
|
especially I guess um with Jane that that would help for since we have this new data now uh in order to go from the transcripts more easily to um just the words that the recognizer would use for scoring. I had to deal with some of it by hand but I think a lot of it can be automated s by
|
Oh one thing I guess I didn't get so you know the language model was straight from from bigram from Switchboard the acoustic models were also from Switchboard or or
|
Yeah.
| false |
QMSum_225
|
especially I guess um with Jane that that would help for since we have this new data now uh in order to go from the transcripts more easily to um just the words that the recognizer would use for scoring. I had to deal with some of it by hand but I think a lot of it can be automated s by
|
Oh one thing I guess I didn't get so you know the language model was straight from from bigram from Switchboard the acoustic models were also from Switchboard or or
|
Yeah.
|
Yeah.
| false |
QMSum_225
|
Oh one thing I guess I didn't get so you know the language model was straight from from bigram from Switchboard the acoustic models were also from Switchboard or or
|
Yeah.
|
Yeah.
|
So they didn't have anything from this acoustic data in yet ?
| false |
QMSum_225
|
Yeah.
|
Yeah.
|
So they didn't have anything from this acoustic data in yet ?
|
That 's amazing.
| false |
QMSum_225
|
Yeah.
|
So they didn't have anything from this acoustic data in yet ?
|
That 's amazing.
|
Yeah , so that 's great.
| false |
QMSum_225
|
So they didn't have anything from this acoustic data in yet ?
|
That 's amazing.
|
Yeah , so that 's great.
|
No.
| false |
QMSum_225
|
That 's amazing.
|
Yeah , so that 's great.
|
No.
|
OK.
| false |
QMSum_225
|
Yeah , so that 's great.
|
No.
|
OK.
|
And actually we actually um used Switchboard telephone bandwidth models
| false |
QMSum_225
|
No.
|
OK.
|
And actually we actually um used Switchboard telephone bandwidth models
|
That 's amazing.
| false |
QMSum_225
|
OK.
|
And actually we actually um used Switchboard telephone bandwidth models
|
That 's amazing.
|
Well that 's those are the only we ones there are ,
| false |
QMSum_225
|
And actually we actually um used Switchboard telephone bandwidth models
|
That 's amazing.
|
Well that 's those are the only we ones there are ,
|
Yeah.
| false |
QMSum_225
|
That 's amazing.
|
Well that 's those are the only we ones there are ,
|
Yeah.
|
which I guess
| false |
QMSum_225
|
Well that 's those are the only we ones there are ,
|
Yeah.
|
which I guess
|
I was just gonna say ,
| false |
QMSum_225
|
Yeah.
|
which I guess
|
I was just gonna say ,
|
so that 's the on that 's the only acoustic training data that we have a lot of
| false |
QMSum_225
|
which I guess
|
I was just gonna say ,
|
so that 's the on that 's the only acoustic training data that we have a lot of
|
yeah.
| false |
QMSum_225
|
I was just gonna say ,
|
so that 's the on that 's the only acoustic training data that we have a lot of
|
yeah.
|
I mean
| false |
QMSum_225
|
so that 's the on that 's the only acoustic training data that we have a lot of
|
yeah.
|
I mean
|
Yeah.
| false |
QMSum_225
|
yeah.
|
I mean
|
Yeah.
|
Right.
| false |
QMSum_225
|
I mean
|
Yeah.
|
Right.
|
and I guess Ramana , so a guy at SRI said that um there 's not a huge amount of difference going from
| false |
QMSum_225
|
Yeah.
|
Right.
|
and I guess Ramana , so a guy at SRI said that um there 's not a huge amount of difference going from
|
Right.
| false |
QMSum_225
|
Right.
|
and I guess Ramana , so a guy at SRI said that um there 's not a huge amount of difference going from
|
Right.
|
it 's it 's not like we probably lose a huge amount but we won't know because we don't have any full band models for s conversational speech.
| false |
QMSum_225
|
and I guess Ramana , so a guy at SRI said that um there 's not a huge amount of difference going from
|
Right.
|
it 's it 's not like we probably lose a huge amount but we won't know because we don't have any full band models for s conversational speech.
|
It 's probably not as bad as going f using full band models on telephone band speech
| false |
QMSum_225
|
Right.
|
it 's it 's not like we probably lose a huge amount but we won't know because we don't have any full band models for s conversational speech.
|
It 's probably not as bad as going f using full band models on telephone band speech
|
So.
| false |
QMSum_225
|
it 's it 's not like we probably lose a huge amount but we won't know because we don't have any full band models for s conversational speech.
|
It 's probably not as bad as going f using full band models on telephone band speech
|
So.
|
Oh yeah.
| false |
QMSum_225
|
It 's probably not as bad as going f using full band models on telephone band speech
|
So.
|
Oh yeah.
|
Right.
| false |
QMSum_225
|
So.
|
Oh yeah.
|
Right.
|
right ?
| false |
QMSum_225
|
Oh yeah.
|
Right.
|
right ?
|
Yeah.
| false |
QMSum_225
|
Right.
|
right ?
|
Yeah.
|
Yeah ,
| false |
QMSum_225
|
right ?
|
Yeah.
|
Yeah ,
|
Right , so it 's so
| false |
QMSum_225
|
Yeah.
|
Yeah ,
|
Right , so it 's so
|
but for Broadcast News when we we played around between the two there wasn't a huge loss.
| false |
QMSum_225
|
Yeah ,
|
Right , so it 's so
|
but for Broadcast News when we we played around between the two there wasn't a huge loss.
|
Right , it was not a big deal.
| false |
QMSum_225
|
Right , so it 's so
|
but for Broadcast News when we we played around between the two there wasn't a huge loss.
|
Right , it was not a big deal.
|
Yeah
| false |
QMSum_225
|
but for Broadcast News when we we played around between the two there wasn't a huge loss.
|
Right , it was not a big deal.
|
Yeah
|
I should I should say that the language model is not just Switchboard
| false |
QMSum_225
|
Right , it was not a big deal.
|
Yeah
|
I should I should say that the language model is not just Switchboard
|
so I wou so that 's good.
| false |
QMSum_225
|
Yeah
|
I should I should say that the language model is not just Switchboard
|
so I wou so that 's good.
|
Although combining em worked well.
| false |
QMSum_225
|
I should I should say that the language model is not just Switchboard
|
so I wou so that 's good.
|
Although combining em worked well.
|
it 's also I mean there 's uh actually more data is from Broadcast News but with a little less weight
| false |
QMSum_225
|
so I wou so that 's good.
|
Although combining em worked well.
|
it 's also I mean there 's uh actually more data is from Broadcast News but with a little less weight
|
Yeah.
| false |
QMSum_225
|
Although combining em worked well.
|
it 's also I mean there 's uh actually more data is from Broadcast News but with a little less weight
|
Yeah.
|
uh because
| false |
QMSum_225
|
it 's also I mean there 's uh actually more data is from Broadcast News but with a little less weight
|
Yeah.
|
uh because
|
Uh - huh.
| false |
QMSum_225
|
Yeah.
|
uh because
|
Uh - huh.
|
Like Trent Lott must have been from
| false |
QMSum_225
|
uh because
|
Uh - huh.
|
Like Trent Lott must have been from
|
mm - hmm , right.
| false |
QMSum_225
|
Uh - huh.
|
Like Trent Lott must have been from
|
mm - hmm , right.
|
I guess Switchboard was before
| false |
QMSum_225
|
Like Trent Lott must have been from
|
mm - hmm , right.
|
I guess Switchboard was before
|
Um By the way just for fun we also ran ,
| false |
QMSum_225
|
mm - hmm , right.
|
I guess Switchboard was before
|
Um By the way just for fun we also ran ,
|
uh.
| false |
QMSum_225
|
I guess Switchboard was before
|
Um By the way just for fun we also ran ,
|
uh.
|
Good point.
| false |
QMSum_225
|
Um By the way just for fun we also ran ,
|
uh.
|
Good point.
|
I mean our complete system starts by doing ge a gender detection
| false |
QMSum_225
|
uh.
|
Good point.
|
I mean our complete system starts by doing ge a gender detection
|
Mm - hmm.
| false |
QMSum_225
|
Good point.
|
I mean our complete system starts by doing ge a gender detection
|
Mm - hmm.
|
so just for the heck of it I ran that
| false |
QMSum_225
|
I mean our complete system starts by doing ge a gender detection
|
Mm - hmm.
|
so just for the heck of it I ran that
|
And it said a hundred percent male ?
| false |
QMSum_225
|
Mm - hmm.
|
so just for the heck of it I ran that
|
And it said a hundred percent male ?
|
um and it might be reassuring for everybody to know that it got all the genders right.
| false |
QMSum_225
|
so just for the heck of it I ran that
|
And it said a hundred percent male ?
|
um and it might be reassuring for everybody to know that it got all the genders right.
|
The j
| false |
QMSum_225
|
And it said a hundred percent male ?
|
um and it might be reassuring for everybody to know that it got all the genders right.
|
The j
|
Yeah so
| false |
QMSum_225
|
um and it might be reassuring for everybody to know that it got all the genders right.
|
The j
|
Yeah so
|
Oh it did ?
| false |
QMSum_225
|
The j
|
Yeah so
|
Oh it did ?
|
Oh that 's I 'm glad.
| false |
QMSum_225
|
Yeah so
|
Oh it did ?
|
Oh that 's I 'm glad.
|
It got all two genders ?
| false |
QMSum_225
|
Oh it did ?
|
Oh that 's I 'm glad.
|
It got all two genders ?
|
Yeah but you know Jane and Adam have you kn about equal performance
| false |
QMSum_225
|
Oh that 's I 'm glad.
|
It got all two genders ?
|
Yeah but you know Jane and Adam have you kn about equal performance
|
Yeah. Yes.
| false |
QMSum_225
|
It got all two genders ?
|
Yeah but you know Jane and Adam have you kn about equal performance
|
Yeah. Yes.
|
and uh and that 's interesting cuz I think the their language models are quite different so and I I 'm pretty sure from listening to Eric that , you know given the words he was saying and given his pronunciation that the reason that he 's so much worse is the lapel.
| false |
QMSum_225
|
Yeah but you know Jane and Adam have you kn about equal performance
|
Yeah. Yes.
|
and uh and that 's interesting cuz I think the their language models are quite different so and I I 'm pretty sure from listening to Eric that , you know given the words he was saying and given his pronunciation that the reason that he 's so much worse is the lapel.
|
Yeah.
| false |
QMSum_225
|
Yeah. Yes.
|
and uh and that 's interesting cuz I think the their language models are quite different so and I I 'm pretty sure from listening to Eric that , you know given the words he was saying and given his pronunciation that the reason that he 's so much worse is the lapel.
|
Yeah.
|
Right.
| false |
QMSum_225
|
and uh and that 's interesting cuz I think the their language models are quite different so and I I 'm pretty sure from listening to Eric that , you know given the words he was saying and given his pronunciation that the reason that he 's so much worse is the lapel.
|
Yeah.
|
Right.
|
That makes a lot of sense ,
| false |
QMSum_225
|
Yeah.
|
Right.
|
That makes a lot of sense ,
|
So it 's nice now if we can just sort of eliminate the lapel one when when we get new microphones
| false |
QMSum_225
|
Right.
|
That makes a lot of sense ,
|
So it 's nice now if we can just sort of eliminate the lapel one when when we get new microphones
|
yeah. Very possible.
| false |
QMSum_225
|
That makes a lot of sense ,
|
So it 's nice now if we can just sort of eliminate the lapel one when when we get new microphones
|
yeah. Very possible.
|
Yeah I I I would bet on that too
| false |
QMSum_225
|
So it 's nice now if we can just sort of eliminate the lapel one when when we get new microphones
|
yeah. Very possible.
|
Yeah I I I would bet on that too
|
that would be worth it
| false |
QMSum_225
|
yeah. Very possible.
|
Yeah I I I would bet on that too
|
that would be worth it
|
cuz he certainly in that when as a as a burp user he was he was a pretty uh strong one.
| false |
QMSum_225
|
Yeah I I I would bet on that too
|
that would be worth it
|
cuz he certainly in that when as a as a burp user he was he was a pretty uh strong one.
|
um Yeah
| false |
QMSum_225
|
that would be worth it
|
cuz he certainly in that when as a as a burp user he was he was a pretty uh strong one.
|
um Yeah
|
Sheep.
| false |
QMSum_225
|
cuz he certainly in that when as a as a burp user he was he was a pretty uh strong one.
|
um Yeah
|
Sheep.
|
he he he sounded to me just from he sounded like a ,
| false |
QMSum_225
|
um Yeah
|
Sheep.
|
he he he sounded to me just from he sounded like a ,
|
Yeah.
| false |
QMSum_225
|
Sheep.
|
he he he sounded to me just from he sounded like a ,
|
Yeah.
|
what 's it a sheep or a goat ?
| false |
QMSum_225
|
he he he sounded to me just from he sounded like a ,
|
Yeah.
|
what 's it a sheep or a goat ?
|
Sheep.
| false |
QMSum_225
|
Yeah.
|
what 's it a sheep or a goat ?
|
Sheep.
|
A sheep.
| false |
QMSum_225
|
what 's it a sheep or a goat ?
|
Sheep.
|
A sheep.
|
Sheep ,
| false |
QMSum_225
|
Sheep.
|
A sheep.
|
Sheep ,
|
Baah.
| false |
QMSum_225
|
A sheep.
|
Sheep ,
|
Baah.
|
Yeah. Sheep is good.
| false |
QMSum_225
|
Sheep ,
|
Baah.
|
Yeah. Sheep is good.
|
right. Sounded good.
| false |
QMSum_225
|
Baah.
|
Yeah. Sheep is good.
|
right. Sounded good.
|
Yeah.
| false |
QMSum_225
|
Yeah. Sheep is good.
|
right. Sounded good.
|
Yeah.
|
Right so um so I guess the good news is that
| false |
QMSum_225
|
right. Sounded good.
|
Yeah.
|
Right so um so I guess the good news is that
|
Mm - hmm.
| false |
QMSum_225
|
Yeah.
|
Right so um so I guess the good news is that
|
Mm - hmm.
|
and and again this is without a lot of the sort of bells and whistles that we c can do with the SRI system and we 'll have more data and we can also start to maybe adapt the language models once we have enough meetings. So this is only twenty minutes of one meeting with no no tailoring at all.
| false |
QMSum_225
|
Right so um so I guess the good news is that
|
Mm - hmm.
|
and and again this is without a lot of the sort of bells and whistles that we c can do with the SRI system and we 'll have more data and we can also start to maybe adapt the language models once we have enough meetings. So this is only twenty minutes of one meeting with no no tailoring at all.
|
I mean clearly there are um with just a small amount of uh actual meeting transcriptions uh thrown into the language model you can probably do quite a bit better because the
| false |
QMSum_225
|
Mm - hmm.
|
and and again this is without a lot of the sort of bells and whistles that we c can do with the SRI system and we 'll have more data and we can also start to maybe adapt the language models once we have enough meetings. So this is only twenty minutes of one meeting with no no tailoring at all.
|
I mean clearly there are um with just a small amount of uh actual meeting transcriptions uh thrown into the language model you can probably do quite a bit better because the
|
Yeah. The voca the vocabulary especially
| false |
QMSum_225
|
and and again this is without a lot of the sort of bells and whistles that we c can do with the SRI system and we 'll have more data and we can also start to maybe adapt the language models once we have enough meetings. So this is only twenty minutes of one meeting with no no tailoring at all.
|
I mean clearly there are um with just a small amount of uh actual meeting transcriptions uh thrown into the language model you can probably do quite a bit better because the
|
Yeah. The voca the vocabulary especially
|
Or just dictionary.
| false |
QMSum_225
|
I mean clearly there are um with just a small amount of uh actual meeting transcriptions uh thrown into the language model you can probably do quite a bit better because the
|
Yeah. The voca the vocabulary especially
|
Or just dictionary.
|
yeah.
| false |
QMSum_225
|
Yeah. The voca the vocabulary especially
|
Or just dictionary.
|
yeah.
|
Not that much the vocabulary actually
| false |
QMSum_225
|
Or just dictionary.
|
yeah.
|
Not that much the vocabulary actually
|
Yeah , so.
| false |
QMSum_225
|
yeah.
|
Not that much the vocabulary actually
|
Yeah , so.
|
I think um well we have to see but it 's uh
| false |
QMSum_225
|
Not that much the vocabulary actually
|
Yeah , so.
|
I think um well we have to see but it 's uh
|
Yeah. It 's pretty good um so then
| false |
QMSum_225
|
Yeah , so.
|
I think um well we have to see but it 's uh
|
Yeah. It 's pretty good um so then
|
Have to add PZM and so on
| false |
QMSum_225
|
I think um well we have to see but it 's uh
|
Yeah. It 's pretty good um so then
|
Have to add PZM and so on
|
And I have to try it on the far field mike
| false |
QMSum_225
|
Yeah. It 's pretty good um so then
|
Have to add PZM and so on
|
And I have to try it on the far field mike
|
but
| false |
QMSum_225
|
Have to add PZM and so on
|
And I have to try it on the far field mike
|
but
|
PZM
| false |
QMSum_225
|
And I have to try it on the far field mike
|
but
|
PZM
|
yeah.
| false |
QMSum_225
|
but
|
PZM
|
yeah.
|
and then there 's things like for the transcription I got when someone has a digit in the transcript I don't know if they said , you know one one or eleven and I don't know if they said Tcl or TCL. there 's things like that where , you know the um we 'll probably have to ask the transcribers to indicate some of those kinds of things but in general it was really good and I 'm hoping and this is this is good news because that means the force alignments should be good and if the force alignments , I mean it 's good news anyway but if the force alignments are good we can get all kinds of information. For example about , you know prosodic information and speaker overlaps and so forth directly from the aligned times. Um so that 'll be something that actually in order to assess the forced alignment um we need s some linguists or some people to look at it and say are these boundaries in about the right place. Because it 's just gonna give us time marks
| false |
QMSum_225
|
PZM
|
yeah.
|
and then there 's things like for the transcription I got when someone has a digit in the transcript I don't know if they said , you know one one or eleven and I don't know if they said Tcl or TCL. there 's things like that where , you know the um we 'll probably have to ask the transcribers to indicate some of those kinds of things but in general it was really good and I 'm hoping and this is this is good news because that means the force alignments should be good and if the force alignments , I mean it 's good news anyway but if the force alignments are good we can get all kinds of information. For example about , you know prosodic information and speaker overlaps and so forth directly from the aligned times. Um so that 'll be something that actually in order to assess the forced alignment um we need s some linguists or some people to look at it and say are these boundaries in about the right place. Because it 's just gonna give us time marks
|
But you know
| false |
QMSum_225
|
yeah.
|
and then there 's things like for the transcription I got when someone has a digit in the transcript I don't know if they said , you know one one or eleven and I don't know if they said Tcl or TCL. there 's things like that where , you know the um we 'll probably have to ask the transcribers to indicate some of those kinds of things but in general it was really good and I 'm hoping and this is this is good news because that means the force alignments should be good and if the force alignments , I mean it 's good news anyway but if the force alignments are good we can get all kinds of information. For example about , you know prosodic information and speaker overlaps and so forth directly from the aligned times. Um so that 'll be something that actually in order to assess the forced alignment um we need s some linguists or some people to look at it and say are these boundaries in about the right place. Because it 's just gonna give us time marks
|
But you know
|
Well we 've done that for one meeting.
| false |
QMSum_225
|
and then there 's things like for the transcription I got when someone has a digit in the transcript I don't know if they said , you know one one or eleven and I don't know if they said Tcl or TCL. there 's things like that where , you know the um we 'll probably have to ask the transcribers to indicate some of those kinds of things but in general it was really good and I 'm hoping and this is this is good news because that means the force alignments should be good and if the force alignments , I mean it 's good news anyway but if the force alignments are good we can get all kinds of information. For example about , you know prosodic information and speaker overlaps and so forth directly from the aligned times. Um so that 'll be something that actually in order to assess the forced alignment um we need s some linguists or some people to look at it and say are these boundaries in about the right place. Because it 's just gonna give us time marks
|
But you know
|
Well we 've done that for one meeting.
|
so. For forced alignment.
| false |
QMSum_225
|
But you know
|
Well we 've done that for one meeting.
|
so. For forced alignment.
|
Uh oh oh f not for words
| false |
QMSum_225
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.