Datasets:
elplaguister
/
DTS_line_datasets

Dataset card Data Studio Files Community
topic_shift
bool
2 classes
utterance
stringlengths
1
7.9k
session_id
stringlengths
7
14
false
I remember when I was when I I I I used , uh , um , a prominent laboratory 's , uh , uh , speech recognizer about , uh This was , boy , this was a while ago , this was about twelve twelve years ago or something. And , um , they were they were perturbed with me because I was breathing in instead of breathing out. And they had models for they they had Markov models for br breathing out but they didn't have them for breathing in.
QMSum_78
false
Yeah.
QMSum_78
false
Uh
QMSum_78
false
That 's interesting. Well , what I wondered is whether it 's possible to have to maybe use the display at the beginning
QMSum_78
false
Yeah.
QMSum_78
false
to be able to to judge how how correctly I mean , have someone do some routine whatever , and and then see if when they 're breathing it 's showing.
QMSum_78
false
I mean , when when it 's on , you can see it.
QMSum_78
false
I don't know if the if it 's
QMSum_78
false
You can definitely see it.
QMSum_78
false
Can you see the breathing ?
QMSum_78
false
Absolutely.
QMSum_78
false
Cuz I
QMSum_78
false
Absolutely.
QMSum_78
false
Oh.
QMSum_78
false
Yeah.
QMSum_78
false
And so , you know , I 've I 've sat here and watched sometimes the breathing ,
QMSum_78
false
and the bar going up and down , and I 'm thinking , I could say something , but
QMSum_78
false
I mean , I think
QMSum_78
false
I don't want to make people self - conscious. Stop breathing !
QMSum_78
false
It it 's going to be imperfect.
QMSum_78
false
Yeah. Uh - huh.
QMSum_78
false
You 're not gonna get it perfect. And you can do some , uh , you know , first - order thing about it , which is to have people move it , uh , uh , a away from being just directly in front of the middle
QMSum_78
false
Yeah.
QMSum_78
false
Good.
QMSum_78
false
but not too far away.
QMSum_78
false
Yeah , i
QMSum_78
false
And then , you know , I think there 's not much Because you can't al you know , interfere w you can't fine tune the meeting that much , I think.
QMSum_78
false
Right.
QMSum_78
false
Yeah.
QMSum_78
false
It 's sort of
QMSum_78
false
That 's true. It just seems like i if something l simple like that can be tweaked and the quality goes , you know , uh , dramatically up , then it might be worth doing.
QMSum_78
false
Yep. And then also the position of the mike also. If it 's more directly , you 'll get better volume. So so , like , yours is pretty far down below your mouth. Yeah.
QMSum_78
false
Yeah.
QMSum_78
false
But Mm - hmm. My my feedback from the transcribers is he is always close to crystal clear and and just fan fantastic to
QMSum_78
false
Yeah.
QMSum_78
false
Mmm , yeah.
QMSum_78
false
Mm - hmm.
QMSum_78
false
I don't know why that is.
QMSum_78
false
Well , I mean , you Yeah , of course. You 're you 're also uh , your volume is is greater. But but still , I mean , they they say
QMSum_78
false
I 've been eating a lot.
QMSum_78
false
I it makes their their job extremely easy.
QMSum_78
false
Uh.
QMSum_78
false
Yeah.
QMSum_78
false
And then there 's mass.
QMSum_78
false
Mm - hmm.
QMSum_78
false
Anyway.
QMSum_78
false
I could say something about about the Well , I don't know what you wanna do. Yeah.
QMSum_78
false
About what ?
QMSum_78
false
About the transcribers or anything or ? I don't know.
QMSum_78
false
Well , the other
QMSum_78
false
But , uh , just to to , um
QMSum_78
false
why don't we do that ?
QMSum_78
false
One more remark , uh , concerning the SRI recognizer. Um. It is useful to transcribe and then ultimately train models for things like breath , and also laughter is very , very frequent and important to to model.
QMSum_78
false
Mm - hmm.
QMSum_78
false
So ,
QMSum_78
false
So ,
QMSum_78
false
if you can in your transcripts mark
QMSum_78
false
mark them ?
QMSum_78
false
mark very audible breaths and laughter especially ,
QMSum_78
false
Mmm.
QMSum_78
false
um
QMSum_78
false
They are.
QMSum_78
false
OK.
QMSum_78
false
They 're putting Eh , so in curly brackets they put " inhale " or " breath ".
QMSum_78
false
Oh , great.
QMSum_78
false
Mm - hmm.
QMSum_78
false
It they and then in curly brackets they say " laughter ". Now they 're they 're not being awfully precise , uh , m So they 're two types of laughter that are not being distinguished.
QMSum_78
false
Mm - hmm.
QMSum_78
false
One is when sometimes s someone will start laughing when they 're in the middle of a sentence.
QMSum_78
false
Mm - hmm.
QMSum_78
false
And and then the other one is when they finish the sentence and then they laugh. So , um , I I did s I did some double checking to look through I mean , you 'd need to have extra e extra complications , like time tags indicating the beginning and ending of of the laughing through the utterance.
QMSum_78
false
It 's not so I don't think it 's , um
QMSum_78
false
And that and what they 're doing is in both cases just saying " curly brackets laughing " a after the unit.
QMSum_78
false
As as long as there is an indication that there was laughter somewhere between two words I think that 's sufficient ,
QMSum_78
false
Yeah.
QMSum_78
false
Good. Oh !
QMSum_78
false
Against they could do forced alignment.
QMSum_78
false
OK.
QMSum_78
false
because actually the recognition of laughter once you kn um you know , is pretty good.
QMSum_78
false
Yeah.
QMSum_78
false
So as long as you can stick a you know , a t a tag in there that that indicates that there was laughter ,
QMSum_78
false
Oh , I didn't know that.
QMSum_78
false
that would probably be , uh , sufficient to train models.
QMSum_78
false
OK.
QMSum_78
false
That would be a really interesting prosodic feature ,
QMSum_78
false
Then
QMSum_78
false
Yeah.
QMSum_78
false
And let me ask y and I gotta ask you one thing about that.
QMSum_78
false
when
QMSum_78
false
Hmm.
QMSum_78
false
So , um , if they laugh between two words , you you 'd get it in between the two words.
QMSum_78
false
Mm - hmm. Right.
QMSum_78
false
But if they laugh across three or four words you you get it after those four words. Does that matter ?
QMSum_78
false
Yeah.
QMSum_78
false
Well , the thing that you is hard to deal with is whe when they speak while laughing. Um , and that 's , uh I don't think that we can do very well with that.
QMSum_78
false
Right.
QMSum_78
false
So
QMSum_78
false
Yeah.
QMSum_78
false
But , um , that 's not as frequent as just laughing between speaking ,
QMSum_78
false
OK.
QMSum_78