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
|
|---|---|---|---|---|---|
For the frame I guess. I computed two - twenty Yeah , well , it 's I guess it 's for the fr the
|
OK. And then there 's delta besides that ?
|
So this is the features that are used by our network and then afterwards , you have to compute the delta on the , uh , main feature stream ,
|
OK.
| false |
QMSum_194
|
OK. And then there 's delta besides that ?
|
So this is the features that are used by our network and then afterwards , you have to compute the delta on the , uh , main feature stream ,
|
OK.
|
which is um , delta and double - deltas , which is fifty milliseconds.
| false |
QMSum_194
|
So this is the features that are used by our network and then afterwards , you have to compute the delta on the , uh , main feature stream ,
|
OK.
|
which is um , delta and double - deltas , which is fifty milliseconds.
|
Yeah. No , I mean , the after the noise part , the forty the the other hundred and eighty Well , I mean , Wait a minute. Some of this is , uh is , uh is in parallel , isn't it ? I mean , the LDA Oh , you have the LDA as part of the V D - uh , VAD ? Or
| false |
QMSum_194
|
OK.
|
which is um , delta and double - deltas , which is fifty milliseconds.
|
Yeah. No , I mean , the after the noise part , the forty the the other hundred and eighty Well , I mean , Wait a minute. Some of this is , uh is , uh is in parallel , isn't it ? I mean , the LDA Oh , you have the LDA as part of the V D - uh , VAD ? Or
|
The VAD use , uh , LDA filtered features also.
| false |
QMSum_194
|
which is um , delta and double - deltas , which is fifty milliseconds.
|
Yeah. No , I mean , the after the noise part , the forty the the other hundred and eighty Well , I mean , Wait a minute. Some of this is , uh is , uh is in parallel , isn't it ? I mean , the LDA Oh , you have the LDA as part of the V D - uh , VAD ? Or
|
The VAD use , uh , LDA filtered features also.
|
Oh , it does ?
| false |
QMSum_194
|
Yeah. No , I mean , the after the noise part , the forty the the other hundred and eighty Well , I mean , Wait a minute. Some of this is , uh is , uh is in parallel , isn't it ? I mean , the LDA Oh , you have the LDA as part of the V D - uh , VAD ? Or
|
The VAD use , uh , LDA filtered features also.
|
Oh , it does ?
|
Mm - hmm.
| false |
QMSum_194
|
The VAD use , uh , LDA filtered features also.
|
Oh , it does ?
|
Mm - hmm.
|
Ah. So in that case there isn't too much in parallel. Uh
| false |
QMSum_194
|
Oh , it does ?
|
Mm - hmm.
|
Ah. So in that case there isn't too much in parallel. Uh
|
No. There is , um , just downsampling , upsampling , and the LDA.
| false |
QMSum_194
|
Mm - hmm.
|
Ah. So in that case there isn't too much in parallel. Uh
|
No. There is , um , just downsampling , upsampling , and the LDA.
|
Um , so the delta at the end is how much ?
| false |
QMSum_194
|
Ah. So in that case there isn't too much in parallel. Uh
|
No. There is , um , just downsampling , upsampling , and the LDA.
|
Um , so the delta at the end is how much ?
|
It 's fifty.
| false |
QMSum_194
|
No. There is , um , just downsampling , upsampling , and the LDA.
|
Um , so the delta at the end is how much ?
|
It 's fifty.
|
It 's
| false |
QMSum_194
|
Um , so the delta at the end is how much ?
|
It 's fifty.
|
It 's
|
Fifty. Alright. So
| false |
QMSum_194
|
It 's fifty.
|
It 's
|
Fifty. Alright. So
|
But well , we could probably put the delta , um , before on - line normalization. It should not that make a big difference ,
| false |
QMSum_194
|
It 's
|
Fifty. Alright. So
|
But well , we could probably put the delta , um , before on - line normalization. It should not that make a big difference ,
|
What if you used a smaller window for the delta ?
| false |
QMSum_194
|
Fifty. Alright. So
|
But well , we could probably put the delta , um , before on - line normalization. It should not that make a big difference ,
|
What if you used a smaller window for the delta ?
|
because
| false |
QMSum_194
|
But well , we could probably put the delta , um , before on - line normalization. It should not that make a big difference ,
|
What if you used a smaller window for the delta ?
|
because
|
Could that help a little bit ? I mean , I guess there 's a lot of things you could do to
| false |
QMSum_194
|
What if you used a smaller window for the delta ?
|
because
|
Could that help a little bit ? I mean , I guess there 's a lot of things you could do to
|
Yeah.
| false |
QMSum_194
|
because
|
Could that help a little bit ? I mean , I guess there 's a lot of things you could do to
|
Yeah.
|
Yeah.
| false |
QMSum_194
|
Could that help a little bit ? I mean , I guess there 's a lot of things you could do to
|
Yeah.
|
Yeah.
|
Yeah ,
| false |
QMSum_194
|
Yeah.
|
Yeah.
|
Yeah ,
|
So
| false |
QMSum_194
|
Yeah.
|
Yeah ,
|
So
|
but , nnn
| false |
QMSum_194
|
Yeah ,
|
So
|
but , nnn
|
Yeah. So if you if you put the delta before the , uh , ana on - line If Yeah
| false |
QMSum_194
|
So
|
but , nnn
|
Yeah. So if you if you put the delta before the , uh , ana on - line If Yeah
|
Mm - hmm.
| false |
QMSum_194
|
but , nnn
|
Yeah. So if you if you put the delta before the , uh , ana on - line If Yeah
|
Mm - hmm.
|
uh then then it could go in parallel.
| false |
QMSum_194
|
Yeah. So if you if you put the delta before the , uh , ana on - line If Yeah
|
Mm - hmm.
|
uh then then it could go in parallel.
|
Cuz i
| false |
QMSum_194
|
Mm - hmm.
|
uh then then it could go in parallel.
|
Cuz i
|
And then y then you don't have that additive
| false |
QMSum_194
|
uh then then it could go in parallel.
|
Cuz i
|
And then y then you don't have that additive
|
Yeah ,
| false |
QMSum_194
|
Cuz i
|
And then y then you don't have that additive
|
Yeah ,
|
Yep.
| false |
QMSum_194
|
And then y then you don't have that additive
|
Yeah ,
|
Yep.
|
cuz the time constant of the on - line normalization is pretty long compared to the delta window ,
| false |
QMSum_194
|
Yeah ,
|
Yep.
|
cuz the time constant of the on - line normalization is pretty long compared to the delta window ,
|
OK.
| false |
QMSum_194
|
Yep.
|
cuz the time constant of the on - line normalization is pretty long compared to the delta window ,
|
OK.
|
so. It should not make
| false |
QMSum_194
|
cuz the time constant of the on - line normalization is pretty long compared to the delta window ,
|
OK.
|
so. It should not make
|
OK. And you ought to be able to shove tw , uh sh uh pull off twenty milliseconds from somewhere else to get it under two hundred , right ? I mean
| false |
QMSum_194
|
OK.
|
so. It should not make
|
OK. And you ought to be able to shove tw , uh sh uh pull off twenty milliseconds from somewhere else to get it under two hundred , right ? I mean
|
Is two hundred the d
| false |
QMSum_194
|
so. It should not make
|
OK. And you ought to be able to shove tw , uh sh uh pull off twenty milliseconds from somewhere else to get it under two hundred , right ? I mean
|
Is two hundred the d
|
The hundred milla
| false |
QMSum_194
|
OK. And you ought to be able to shove tw , uh sh uh pull off twenty milliseconds from somewhere else to get it under two hundred , right ? I mean
|
Is two hundred the d
|
The hundred milla
|
Mm - hmm.
| false |
QMSum_194
|
Is two hundred the d
|
The hundred milla
|
Mm - hmm.
|
mill a hundred milliseconds for smoothing is sort of an arbitrary amount. It could be eighty and and probably do @ @
| false |
QMSum_194
|
The hundred milla
|
Mm - hmm.
|
mill a hundred milliseconds for smoothing is sort of an arbitrary amount. It could be eighty and and probably do @ @
|
Yeah ,
| false |
QMSum_194
|
Mm - hmm.
|
mill a hundred milliseconds for smoothing is sort of an arbitrary amount. It could be eighty and and probably do @ @
|
Yeah ,
|
i a hun
| false |
QMSum_194
|
mill a hundred milliseconds for smoothing is sort of an arbitrary amount. It could be eighty and and probably do @ @
|
Yeah ,
|
i a hun
|
yeah.
| false |
QMSum_194
|
Yeah ,
|
i a hun
|
yeah.
|
uh Wh - what 's the baseline you need to be under ? Two hundred ?
| false |
QMSum_194
|
i a hun
|
yeah.
|
uh Wh - what 's the baseline you need to be under ? Two hundred ?
|
Well , we don't know. They 're still arguing about it.
| false |
QMSum_194
|
yeah.
|
uh Wh - what 's the baseline you need to be under ? Two hundred ?
|
Well , we don't know. They 're still arguing about it.
|
Oh.
| false |
QMSum_194
|
uh Wh - what 's the baseline you need to be under ? Two hundred ?
|
Well , we don't know. They 're still arguing about it.
|
Oh.
|
I mean , if it 's two if if it 's , uh if it 's two - fifty , then we could keep the delta where it is if we shaved off twenty. If it 's two hundred , if we shaved off twenty , we could we could , uh , meet it by moving the delta back.
| false |
QMSum_194
|
Well , we don't know. They 're still arguing about it.
|
Oh.
|
I mean , if it 's two if if it 's , uh if it 's two - fifty , then we could keep the delta where it is if we shaved off twenty. If it 's two hundred , if we shaved off twenty , we could we could , uh , meet it by moving the delta back.
|
So , how do you know that what you have is too much if they 're still deciding ?
| false |
QMSum_194
|
Oh.
|
I mean , if it 's two if if it 's , uh if it 's two - fifty , then we could keep the delta where it is if we shaved off twenty. If it 's two hundred , if we shaved off twenty , we could we could , uh , meet it by moving the delta back.
|
So , how do you know that what you have is too much if they 're still deciding ?
|
Uh , we don't , but it 's just I mean , the main thing is that since that we got burned last time , and you know , by not worrying about it very much , we 're just staying conscious of it.
| true |
QMSum_194
|
I mean , if it 's two if if it 's , uh if it 's two - fifty , then we could keep the delta where it is if we shaved off twenty. If it 's two hundred , if we shaved off twenty , we could we could , uh , meet it by moving the delta back.
|
So , how do you know that what you have is too much if they 're still deciding ?
|
Uh , we don't , but it 's just I mean , the main thing is that since that we got burned last time , and you know , by not worrying about it very much , we 're just staying conscious of it.
|
Uh - huh. Oh , OK , I see.
| false |
QMSum_194
|
So , how do you know that what you have is too much if they 're still deciding ?
|
Uh , we don't , but it 's just I mean , the main thing is that since that we got burned last time , and you know , by not worrying about it very much , we 're just staying conscious of it.
|
Uh - huh. Oh , OK , I see.
|
And so , th I mean , if if if a week before we have to be done someone says , " Well , you have to have fifty milliseconds less than you have now " , it would be pretty frantic around here. So
| false |
QMSum_194
|
Uh , we don't , but it 's just I mean , the main thing is that since that we got burned last time , and you know , by not worrying about it very much , we 're just staying conscious of it.
|
Uh - huh. Oh , OK , I see.
|
And so , th I mean , if if if a week before we have to be done someone says , " Well , you have to have fifty milliseconds less than you have now " , it would be pretty frantic around here. So
|
Ah , OK.
| false |
QMSum_194
|
Uh - huh. Oh , OK , I see.
|
And so , th I mean , if if if a week before we have to be done someone says , " Well , you have to have fifty milliseconds less than you have now " , it would be pretty frantic around here. So
|
Ah , OK.
|
Uh
| false |
QMSum_194
|
And so , th I mean , if if if a week before we have to be done someone says , " Well , you have to have fifty milliseconds less than you have now " , it would be pretty frantic around here. So
|
Ah , OK.
|
Uh
|
But still , that 's that 's a pretty big , uh , win. And it doesn't seem like you 're in terms of your delay , you 're , uh , that
| false |
QMSum_194
|
Ah , OK.
|
Uh
|
But still , that 's that 's a pretty big , uh , win. And it doesn't seem like you 're in terms of your delay , you 're , uh , that
|
He added a bit on , I guess , because before we were we were had were able to have the noise , uh , stuff , uh , and the LVA be in parallel.
| false |
QMSum_194
|
Uh
|
But still , that 's that 's a pretty big , uh , win. And it doesn't seem like you 're in terms of your delay , you 're , uh , that
|
He added a bit on , I guess , because before we were we were had were able to have the noise , uh , stuff , uh , and the LVA be in parallel.
|
Hmm.
| false |
QMSum_194
|
But still , that 's that 's a pretty big , uh , win. And it doesn't seem like you 're in terms of your delay , you 're , uh , that
|
He added a bit on , I guess , because before we were we were had were able to have the noise , uh , stuff , uh , and the LVA be in parallel.
|
Hmm.
|
And now he 's he 's requiring it to be done first.
| false |
QMSum_194
|
He added a bit on , I guess , because before we were we were had were able to have the noise , uh , stuff , uh , and the LVA be in parallel.
|
Hmm.
|
And now he 's he 's requiring it to be done first.
|
Well , but I think the main thing , maybe , is the cleaning of the speech , which takes forty milliseconds or so.
| false |
QMSum_194
|
Hmm.
|
And now he 's he 's requiring it to be done first.
|
Well , but I think the main thing , maybe , is the cleaning of the speech , which takes forty milliseconds or so.
|
Right. Well , so you say let 's say ten milliseconds seconds for the LDA.
| false |
QMSum_194
|
And now he 's he 's requiring it to be done first.
|
Well , but I think the main thing , maybe , is the cleaning of the speech , which takes forty milliseconds or so.
|
Right. Well , so you say let 's say ten milliseconds seconds for the LDA.
|
And and but the LDA is , well , pretty short right now.
| false |
QMSum_194
|
Well , but I think the main thing , maybe , is the cleaning of the speech , which takes forty milliseconds or so.
|
Right. Well , so you say let 's say ten milliseconds seconds for the LDA.
|
And and but the LDA is , well , pretty short right now.
|
Well , ten. And then forty for the other.
| false |
QMSum_194
|
Right. Well , so you say let 's say ten milliseconds seconds for the LDA.
|
And and but the LDA is , well , pretty short right now.
|
Well , ten. And then forty for the other.
|
Yeah.
| false |
QMSum_194
|
And and but the LDA is , well , pretty short right now.
|
Well , ten. And then forty for the other.
|
Yeah.
|
Yeah , the LDA LDA we don't know , is , like is it very crucial for the features , right ?
| false |
QMSum_194
|
Well , ten. And then forty for the other.
|
Yeah.
|
Yeah , the LDA LDA we don't know , is , like is it very crucial for the features , right ?
|
No. I just This is the first try.
| false |
QMSum_194
|
Yeah.
|
Yeah , the LDA LDA we don't know , is , like is it very crucial for the features , right ?
|
No. I just This is the first try.
|
Yeah.
| false |
QMSum_194
|
Yeah , the LDA LDA we don't know , is , like is it very crucial for the features , right ?
|
No. I just This is the first try.
|
Yeah.
|
Right ,
| false |
QMSum_194
|
No. I just This is the first try.
|
Yeah.
|
Right ,
|
I mean , I maybe the LDA 's not very useful then.
| false |
QMSum_194
|
Yeah.
|
Right ,
|
I mean , I maybe the LDA 's not very useful then.
|
so you could start pulling back ,
| false |
QMSum_194
|
Right ,
|
I mean , I maybe the LDA 's not very useful then.
|
so you could start pulling back ,
|
S s h
| false |
QMSum_194
|
I mean , I maybe the LDA 's not very useful then.
|
so you could start pulling back ,
|
S s h
|
but
| false |
QMSum_194
|
so you could start pulling back ,
|
S s h
|
but
|
Yeah ,
| false |
QMSum_194
|
S s h
|
but
|
Yeah ,
|
But I think you have
| false |
QMSum_194
|
but
|
Yeah ,
|
But I think you have
|
I mean , you have twenty for delta computation which y now you 're sort of doing twice , right ? But yo w were you doing that before ?
| false |
QMSum_194
|
Yeah ,
|
But I think you have
|
I mean , you have twenty for delta computation which y now you 're sort of doing twice , right ? But yo w were you doing that before ?
|
Mmm. Well , in the proposal , um , the input of the VAD network were just three frames , I think.
| false |
QMSum_194
|
But I think you have
|
I mean , you have twenty for delta computation which y now you 're sort of doing twice , right ? But yo w were you doing that before ?
|
Mmm. Well , in the proposal , um , the input of the VAD network were just three frames , I think.
|
On the in the Mm - hmm. Just Yeah , just the static , no delta.
| false |
QMSum_194
|
I mean , you have twenty for delta computation which y now you 're sort of doing twice , right ? But yo w were you doing that before ?
|
Mmm. Well , in the proposal , um , the input of the VAD network were just three frames , I think.
|
On the in the Mm - hmm. Just Yeah , just the static , no delta.
|
Right.
| false |
QMSum_194
|
Mmm. Well , in the proposal , um , the input of the VAD network were just three frames , I think.
|
On the in the Mm - hmm. Just Yeah , just the static , no delta.
|
Right.
|
Uh , static features.
| false |
QMSum_194
|
On the in the Mm - hmm. Just Yeah , just the static , no delta.
|
Right.
|
Uh , static features.
|
So , what you have now is fort uh , forty for the the noise , twenty for the delta , and ten for the LDA. That 's seventy milliseconds of stuff which was formerly in parallel ,
| false |
QMSum_194
|
Right.
|
Uh , static features.
|
So , what you have now is fort uh , forty for the the noise , twenty for the delta , and ten for the LDA. That 's seventy milliseconds of stuff which was formerly in parallel ,
|
right ? So I think ,
| false |
QMSum_194
|
Uh , static features.
|
So , what you have now is fort uh , forty for the the noise , twenty for the delta , and ten for the LDA. That 's seventy milliseconds of stuff which was formerly in parallel ,
|
right ? So I think ,
|
Mm - hmm.
| false |
QMSum_194
|
So , what you have now is fort uh , forty for the the noise , twenty for the delta , and ten for the LDA. That 's seventy milliseconds of stuff which was formerly in parallel ,
|
right ? So I think ,
|
Mm - hmm.
|
you know , that 's that 's the difference as far as the timing , right ?
| false |
QMSum_194
|
right ? So I think ,
|
Mm - hmm.
|
you know , that 's that 's the difference as far as the timing , right ?
|
Yeah.
| false |
QMSum_194
|
Mm - hmm.
|
you know , that 's that 's the difference as far as the timing , right ?
|
Yeah.
|
Um , and you could experiment with cutting various pieces of these back a bit , but I mean , we 're s we 're not we 're not in terrible shape.
| false |
QMSum_194
|
you know , that 's that 's the difference as far as the timing , right ?
|
Yeah.
|
Um , and you could experiment with cutting various pieces of these back a bit , but I mean , we 're s we 're not we 're not in terrible shape.
|
Yeah , that 's what it seems like to me. It 's pretty good.
| false |
QMSum_194
|
Yeah.
|
Um , and you could experiment with cutting various pieces of these back a bit , but I mean , we 're s we 're not we 're not in terrible shape.
|
Yeah , that 's what it seems like to me. It 's pretty good.
|
Yeah.
| false |
QMSum_194
|
Um , and you could experiment with cutting various pieces of these back a bit , but I mean , we 're s we 're not we 're not in terrible shape.
|
Yeah , that 's what it seems like to me. It 's pretty good.
|
Yeah.
|
Mm - hmm.
| false |
QMSum_194
|
Yeah , that 's what it seems like to me. It 's pretty good.
|
Yeah.
|
Mm - hmm.
|
It 's it 's not like it 's adding up to four hundred milliseconds or something.
| false |
QMSum_194
|
Yeah.
|
Mm - hmm.
|
It 's it 's not like it 's adding up to four hundred milliseconds or something.
|
Where where is this where is this fifty - seven point O two in in comparison to the last evaluation ?
| false |
QMSum_194
|
Mm - hmm.
|
It 's it 's not like it 's adding up to four hundred milliseconds or something.
|
Where where is this where is this fifty - seven point O two in in comparison to the last evaluation ?
|
Well , it 's I think it 's better than anything , uh , anybody got.
| false |
QMSum_194
|
It 's it 's not like it 's adding up to four hundred milliseconds or something.
|
Where where is this where is this fifty - seven point O two in in comparison to the last evaluation ?
|
Well , it 's I think it 's better than anything , uh , anybody got.
|
Oh , is that right ?
| false |
QMSum_194
|
Where where is this where is this fifty - seven point O two in in comparison to the last evaluation ?
|
Well , it 's I think it 's better than anything , uh , anybody got.
|
Oh , is that right ?
|
Yeah. The best was fifty - four point five.
| false |
QMSum_194
|
Well , it 's I think it 's better than anything , uh , anybody got.
|
Oh , is that right ?
|
Yeah. The best was fifty - four point five.
|
Yeah.
| false |
QMSum_194
|
Oh , is that right ?
|
Yeah. The best was fifty - four point five.
|
Yeah.
|
Point s
| false |
QMSum_194
|
Yeah. The best was fifty - four point five.
|
Yeah.
|
Point s
|
Oh.
| false |
QMSum_194
|
Yeah.
|
Point s
|
Oh.
|
Yeah. Uh
| false |
QMSum_194
|
Point s
|
Oh.
|
Yeah. Uh
|
And our system was forty - nine , but with the neural network.
| false |
QMSum_194
|
Oh.
|
Yeah. Uh
|
And our system was forty - nine , but with the neural network.
|
Wow. So this is almost ten percent.
| false |
QMSum_194
|
Yeah. Uh
|
And our system was forty - nine , but with the neural network.
|
Wow. So this is almost ten percent.
|
With the f with the neural net. Yeah , and r and
| false |
QMSum_194
|
And our system was forty - nine , but with the neural network.
|
Wow. So this is almost ten percent.
|
With the f with the neural net. Yeah , and r and
|
It would
| false |
QMSum_194
|
Wow. So this is almost ten percent.
|
With the f with the neural net. Yeah , and r and
|
It would
|
Yeah , so this is this is like the first proposal. The proposal - one. It was forty - four , actually.
| false |
QMSum_194
|
With the f with the neural net. Yeah , and r and
|
It would
|
Yeah , so this is this is like the first proposal. The proposal - one. It was forty - four , actually.
|
Yeah. Yeah. And we still don't have the neural net in. So so it 's
| false |
QMSum_194
|
It would
|
Yeah , so this is this is like the first proposal. The proposal - one. It was forty - four , actually.
|
Yeah. Yeah. And we still don't have the neural net in. So so it 's
|
Wow.
| false |
QMSum_194
|
Yeah , so this is this is like the first proposal. The proposal - one. It was forty - four , actually.
|
Yeah. Yeah. And we still don't have the neural net in. So so it 's
|
Wow.
|
You know. So it 's We 're we 're doing better.
| false |
QMSum_194
|
Yeah. Yeah. And we still don't have the neural net in. So so it 's
|
Wow.
|
You know. So it 's We 're we 're doing better.
|
This is this is really good.
| false |
QMSum_194
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.