elplaguister/DTS_line_datasets · Datasets at Hugging Face

topic_shift bool 2 classes	utterance stringlengths 1 7.9k	session_id stringlengths 7 14
false	I think you are. Aren't you ?	QMSum_86
false	They both want their cut.	QMSum_86
false	Hmm.	QMSum_86
false	Hmm. Yeah.	QMSum_86
false	Right ?	QMSum_86
false	Yeah. Yeah. Huh. Canada w Canada wants a cut ?	QMSum_86
false	Mm - hmm.	QMSum_86
false	Have to do So you you have to do two returns ?	QMSum_86
false	Mmm. W uh , for two thousand I did. Yeah.	QMSum_86
false	Oh , oh. Yeah. For tw That 's right , ju	QMSum_86
false	But not for this next year ?	QMSum_86
false	Two thousand. Yeah. Probably not this next year , I guess.	QMSum_86
false	Ye	QMSum_86
false	Yeah.	QMSum_86
false	Um.	QMSum_86
false	Yeah.	QMSum_86
false	Uh , I 'll I 'll still have a bit of Canadian income but it 'll be less complicated because I will not be a considered a resident of Canada anymore , so I won't have to declare my American income on my Canadian return.	QMSum_86
true	OK. Alright. Uh. Barry , do you wanna say something about your stuff here ?	QMSum_86
false	Oh , um. Right. I just , um , continuing looking at , uh , ph uh , phonetic events , and , uh , this Tuesday gonna be , uh , meeting with John Ohala with Chuck to talk some more about these , uh , ph um , phonetic events. Um , came up with , uh , a plan of attack , uh , gonna execute , and um Yeah. It 's that 's pretty much it.	QMSum_86
false	Oh , well. No Um , why don't you say something about what it is ?	QMSum_86
false	Oh , you oh , you want you want details. Hmm. OK.	QMSum_86
false	Well , we 're all gathered here together. I thought we 'd , you know	QMSum_86
false	I was hoping I could wave my hands. Um. So , um. So , once wa I I was thinking getting getting us a set of acoustic events to um , to be able to distinguish between , uh , phones and words and stuff. And um , once we we would figure out a set of these events that can be , you know , um , hand - labeled or or derived , uh , from h the hand - labeled phone targets. Um , we could take these events and , um , do some cheating experiments , um , where we feed , um , these events into an SRI system , um , eh , and evaluate its performance on a Switchboard task. Uh , yeah.	QMSum_86
false	Hey , Barry ? Can you give an example of an event ?	QMSum_86
false	Yeah. Sure. Um , I I can give you an example of twenty - odd events. Um So , he In this paper , um , it 's talking about phoneme recognition using acoustic events. So , things like frication or , uh , nasality.	QMSum_86
false	Whose paper is it ?	QMSum_86
false	Um , this is a paper by Hubener and Cardson Benson Bernds - Berndsen.	QMSum_86
false	Yeah. Huh. From , uh , University of Hamburg and Bielefeld.	QMSum_86
false	Mm - hmm.	QMSum_86
false	OK.	QMSum_86
false	Um.	QMSum_86
false	Yeah. I think the just to expand a little bit on the idea of acoustic event.	QMSum_86
false	Mm - hmm.	QMSum_86
false	There 's , um in my mind , anyways , there 's a difference between , um , acoustic features and acoustic events. And I think of acoustic features as being , um , things that linguists talk about , like , um	QMSum_86
false	So , stuff that 's not based on data.	QMSum_86
false	Stuff that 's not based on data , necessarily.	QMSum_86
false	Yeah. Oh , OK. Yeah. Yeah , OK.	QMSum_86
false	Right. That 's not based on , you know , acoustic data. So they talk about features for phones , like , uh , its height ,	QMSum_86
false	Yeah.	QMSum_86
false	its tenseness , laxness , things like that ,	QMSum_86
false	Mm - hmm.	QMSum_86
false	which may or may not be all that easy to measure in the acoustic signal. Versus an acoustic event , which is just some something in the acoustic signal that is fairly easy to measure. Um. So it 's , um it 's a little different , in at least in my mind.	QMSum_86
false	I mean , when we did the SPAM work I mean , there we had we had this notion of an , uh , auditory @ @ auditory event.	QMSum_86
false	Good. That 's great.	QMSum_86
false	And , uh , um , called them " avents " , uh , uh , uh , with an A at the front.	QMSum_86
false	Mm - hmm.	QMSum_86
false	Uh. And the the the idea was something that occurred that is important to a bunch of neurons somewhere. So.	QMSum_86
false	Mm - hmm.	QMSum_86
false	Um. A sudden change or a relatively rapid change in some spectral characteristic will will do sort of this. I mean , there 's certainly a bunch of a bunch of places where you know that neurons are gonna fire because something novel has happened. That was that was the main thing that we were focusing on there. But there 's certainly other things beyond what we talked about there that aren't just sort of rapid changes , but	QMSum_86
false	It 's kinda like the difference between top - down and bottom - up.	QMSum_86
false	Yeah.	QMSum_86
false	I think of the acoustic you know , phonetic features as being top - down. You know , you look at the phone and you say this phone is supposed to be you know , have this feature , this feature , and this feature. Whether tha those features show up in the acoustic signal is sort of irrelevant. Whereas , an acoustic event goes the other way. Here 's the signal. Here 's some event.	QMSum_86
false	Mm - hmm.	QMSum_86
false	What ? And then that you know , that may map to this phone sometimes , and sometimes it may not. It just depen maybe depends on the context , things like that.	QMSum_86
false	Mm - hmm.	QMSum_86
false	And so it 's sort of a different way of looking.	QMSum_86
false	Mm - hmm.	QMSum_86
false	Yeah. So. Yeah.	QMSum_86
false	OK.	QMSum_86
false	Mm - hmm. Um Using these these events , um , you know , we can we can perform these these , uh , cheating experiments. See how how how good they are , um , in , um in terms of phoneme recognition or word recognition. And , um and then from that point on , I would , uh , s design robust event detectors , um , in a similar , um , wa spirit that Saul has done w uh , with his graphical models , and this this probabilistic AND - OR model that he uses. Um , eh , try to extend it to , um to account for other other phenomena like , um , CMR co - modulation release. And , um and maybe also investigate ways to to modify the structure of these models , um , in a data - driven way , uh , similar to the way that , uh , Jeff Jeff , uh , Bilmes did his work. Um , and while I 'm I 'm doing these , um , event detectors , you know , I can ma mea measure my progress by comparing , um , the error rates in clean and noisy conditions to something like , uh , neural nets. Um , and So so , once we have these these , uh , event detectors , um , we could put them together and and feed the outputs of the event detectors into into the SRI , um , HMM HMM system , and , um and test it on on Switchboard or , um , maybe even Aurora stuff. And , that 's pretty much the the big picture of of um , the plan.	QMSum_86
false	By the way , um , there 's , uh , a couple people who are gonna be here I forget if I already told you this , but , a couple people who are gonna be here for six months.	QMSum_86
false	Mm - hmm.	QMSum_86
false	Uh uh , there 's a Professor Kollmeier , uh , from Germany who 's , uh , uh , quite big in the , uh , hearing - aid signal - processing area and , um , Michael Kleinschmidt , who 's worked with him , who also looks at auditory properties inspired by various , uh , brain function things.	QMSum_86
false	Hmm.	QMSum_86
false	So , um , um , I think they 'll be interesting to talk to , in this sort of issue as these detectors are are , uh , developing.	QMSum_86
false	Hmm. OK.	QMSum_86
false	So , he looks at interesting interesting things in in the different ways of looking at spectra in order to to get various speech properties out. So.	QMSum_86
false	OK.	QMSum_86
false	OK. Well , short meeting , but that 's OK. And , uh , we might as well do our digits. And like I say , I I encourage you to go ahead and meet , uh , next week with , uh , uh , Hynek. Alright , I 'll I 'll start. It 's , uh , one thirty - five. seventeen OK	QMSum_86
false	Right first time this time. Nu There we go. It's not that complicated , but I get it wrong every time.	QMSum_90
false	Okay so we are just waiting for Matthew .	QMSum_90
false	For Matthew , yep.	QMSum_90
false	Mm. Uh So I suggest we start the meeting uh without Matthew uh	QMSum_90
false	Mm-hmm.	QMSum_90
false	Mm 'kay.	QMSum_90
false	he's uh obviously late for some reason. Good. Um. Today uh we will uh talk about uh conceptual design. I hope uh you both did some uh some work uh concerning a uh conceptual design.	QMSum_90
false	Mm-hmm.	QMSum_90
false	Um this will be the uh agenda for the meeting uh . Uh I will take some minutes uh again.	QMSum_90
false	Mm-hmm.	QMSum_90
false	Um we will have the presentations of y of you different team members ,	QMSum_90
false	Yep.	QMSum_90
false	Yep.	QMSum_90
false	and then try to come to decisions uh about the concepts uh you have presented. So and that uh will uh	QMSum_90
false	Mm-hmm.	QMSum_90
false	we have some uh forty minutes uh to complete this uh.	QMSum_90
false	Mm-hmm.	QMSum_90
false	So um who has the fir do you ha Anna do you have your presentation ready ?	QMSum_90
false	I have a presentation , I'm just making this	QMSum_90
false	Yeah I think yeah the Matthew it is it's important that Matthew yeah is here	QMSum_90
false	Okay. Ah	QMSum_90
false	because it's really a a team uh project with a team	QMSum_90
false	there is Matthew.	QMSum_90
false	and if someone is not here then we cannot	QMSum_90
false	Sorry.	QMSum_90
false	but it's okay it's good.	QMSum_90
false	Okay I'll just email you this file , my presentation.	QMSum_90
false	So. Good. Do presentation ready ?	QMSum_90
false	Mm-hmm I'm just emailing it to you.	QMSum_90
false	Oh okay.	QMSum_90
false	So	QMSum_90

false

I think you are. Aren't you ?

QMSum_86

false

They both want their cut.

QMSum_86

false

Hmm.

QMSum_86

false

Hmm. Yeah.

QMSum_86

false

Right ?

QMSum_86

false

Yeah. Yeah. Huh. Canada w Canada wants a cut ?

QMSum_86

false

Mm - hmm.

QMSum_86

false

Have to do So you you have to do two returns ?

QMSum_86

false

Mmm. W uh , for two thousand I did. Yeah.

QMSum_86

false

Oh , oh. Yeah. For tw That 's right , ju

QMSum_86

false

But not for this next year ?

QMSum_86

false

Two thousand. Yeah. Probably not this next year , I guess.

QMSum_86

false

Ye

QMSum_86

false

Yeah.

QMSum_86

false

Um.

QMSum_86

false

Yeah.

QMSum_86

false

Uh , I 'll I 'll still have a bit of Canadian income but it 'll be less complicated because I will not be a considered a resident of Canada anymore , so I won't have to declare my American income on my Canadian return.

QMSum_86

true

OK. Alright. Uh. Barry , do you wanna say something about your stuff here ?

QMSum_86

false

Oh , um. Right. I just , um , continuing looking at , uh , ph uh , phonetic events , and , uh , this Tuesday gonna be , uh , meeting with John Ohala with Chuck to talk some more about these , uh , ph um , phonetic events. Um , came up with , uh , a plan of attack , uh , gonna execute , and um Yeah. It 's that 's pretty much it.

QMSum_86

false

Oh , well. No Um , why don't you say something about what it is ?

QMSum_86

false

Oh , you oh , you want you want details. Hmm. OK.

QMSum_86

false

Well , we 're all gathered here together. I thought we 'd , you know

QMSum_86

false

I was hoping I could wave my hands. Um. So , um. So , once wa I I was thinking getting getting us a set of acoustic events to um , to be able to distinguish between , uh , phones and words and stuff. And um , once we we would figure out a set of these events that can be , you know , um , hand - labeled or or derived , uh , from h the hand - labeled phone targets. Um , we could take these events and , um , do some cheating experiments , um , where we feed , um , these events into an SRI system , um , eh , and evaluate its performance on a Switchboard task. Uh , yeah.

QMSum_86

false

Hey , Barry ? Can you give an example of an event ?

QMSum_86

false

Yeah. Sure. Um , I I can give you an example of twenty - odd events. Um So , he In this paper , um , it 's talking about phoneme recognition using acoustic events. So , things like frication or , uh , nasality.

QMSum_86

false

Whose paper is it ?

QMSum_86

false

Um , this is a paper by Hubener and Cardson Benson Bernds - Berndsen.

QMSum_86

false

Yeah. Huh. From , uh , University of Hamburg and Bielefeld.

QMSum_86

false

Mm - hmm.

QMSum_86

false

OK.

QMSum_86

false

Um.

QMSum_86

false

Yeah. I think the just to expand a little bit on the idea of acoustic event.

QMSum_86

false

Mm - hmm.

QMSum_86

false

There 's , um in my mind , anyways , there 's a difference between , um , acoustic features and acoustic events. And I think of acoustic features as being , um , things that linguists talk about , like , um

QMSum_86

false

So , stuff that 's not based on data.

QMSum_86

false

Stuff that 's not based on data , necessarily.

QMSum_86

false

Yeah. Oh , OK. Yeah. Yeah , OK.

QMSum_86

false

Right. That 's not based on , you know , acoustic data. So they talk about features for phones , like , uh , its height ,

QMSum_86

false

Yeah.

QMSum_86

false

its tenseness , laxness , things like that ,

QMSum_86

false

Mm - hmm.

QMSum_86

false

which may or may not be all that easy to measure in the acoustic signal. Versus an acoustic event , which is just some something in the acoustic signal that is fairly easy to measure. Um. So it 's , um it 's a little different , in at least in my mind.

QMSum_86

false

I mean , when we did the SPAM work I mean , there we had we had this notion of an , uh , auditory @ @ auditory event.

QMSum_86

false

Good. That 's great.

QMSum_86

false

And , uh , um , called them " avents " , uh , uh , uh , with an A at the front.

QMSum_86

false

Mm - hmm.

QMSum_86

false

Uh. And the the the idea was something that occurred that is important to a bunch of neurons somewhere. So.

QMSum_86

false

Mm - hmm.

QMSum_86

false

Um. A sudden change or a relatively rapid change in some spectral characteristic will will do sort of this. I mean , there 's certainly a bunch of a bunch of places where you know that neurons are gonna fire because something novel has happened. That was that was the main thing that we were focusing on there. But there 's certainly other things beyond what we talked about there that aren't just sort of rapid changes , but

QMSum_86

false

It 's kinda like the difference between top - down and bottom - up.

QMSum_86

false

Yeah.

QMSum_86

false

I think of the acoustic you know , phonetic features as being top - down. You know , you look at the phone and you say this phone is supposed to be you know , have this feature , this feature , and this feature. Whether tha those features show up in the acoustic signal is sort of irrelevant. Whereas , an acoustic event goes the other way. Here 's the signal. Here 's some event.

QMSum_86

false

Mm - hmm.

QMSum_86

false

What ? And then that you know , that may map to this phone sometimes , and sometimes it may not. It just depen maybe depends on the context , things like that.

QMSum_86

false

Mm - hmm.

QMSum_86

false

And so it 's sort of a different way of looking.

QMSum_86

false

Mm - hmm.

QMSum_86

false

Yeah. So. Yeah.

QMSum_86

false

OK.

QMSum_86

false

Mm - hmm. Um Using these these events , um , you know , we can we can perform these these , uh , cheating experiments. See how how how good they are , um , in , um in terms of phoneme recognition or word recognition. And , um and then from that point on , I would , uh , s design robust event detectors , um , in a similar , um , wa spirit that Saul has done w uh , with his graphical models , and this this probabilistic AND - OR model that he uses. Um , eh , try to extend it to , um to account for other other phenomena like , um , CMR co - modulation release. And , um and maybe also investigate ways to to modify the structure of these models , um , in a data - driven way , uh , similar to the way that , uh , Jeff Jeff , uh , Bilmes did his work. Um , and while I 'm I 'm doing these , um , event detectors , you know , I can ma mea measure my progress by comparing , um , the error rates in clean and noisy conditions to something like , uh , neural nets. Um , and So so , once we have these these , uh , event detectors , um , we could put them together and and feed the outputs of the event detectors into into the SRI , um , HMM HMM system , and , um and test it on on Switchboard or , um , maybe even Aurora stuff. And , that 's pretty much the the big picture of of um , the plan.

QMSum_86

false

By the way , um , there 's , uh , a couple people who are gonna be here I forget if I already told you this , but , a couple people who are gonna be here for six months.

QMSum_86

false

Mm - hmm.

QMSum_86

false

Uh uh , there 's a Professor Kollmeier , uh , from Germany who 's , uh , uh , quite big in the , uh , hearing - aid signal - processing area and , um , Michael Kleinschmidt , who 's worked with him , who also looks at auditory properties inspired by various , uh , brain function things.

QMSum_86

false

Hmm.

QMSum_86

false

So , um , um , I think they 'll be interesting to talk to , in this sort of issue as these detectors are are , uh , developing.

QMSum_86

false

Hmm. OK.

QMSum_86

false

So , he looks at interesting interesting things in in the different ways of looking at spectra in order to to get various speech properties out. So.

QMSum_86

false

OK.

QMSum_86

false

OK. Well , short meeting , but that 's OK. And , uh , we might as well do our digits. And like I say , I I encourage you to go ahead and meet , uh , next week with , uh , uh , Hynek. Alright , I 'll I 'll start. It 's , uh , one thirty - five. seventeen OK

QMSum_86

false

Right first time this time. Nu There we go. It's not that complicated , but I get it wrong every time.

QMSum_90

false

Okay so we are just waiting for Matthew .

QMSum_90

false

For Matthew , yep.

QMSum_90

false

Mm. Uh So I suggest we start the meeting uh without Matthew uh

QMSum_90

false

Mm-hmm.

QMSum_90

false

Mm 'kay.

QMSum_90

false

he's uh obviously late for some reason. Good. Um. Today uh we will uh talk about uh conceptual design. I hope uh you both did some uh some work uh concerning a uh conceptual design.

QMSum_90

false

Mm-hmm.

QMSum_90

false

Um this will be the uh agenda for the meeting uh . Uh I will take some minutes uh again.

QMSum_90

false

Mm-hmm.

QMSum_90

false

Um we will have the presentations of y of you different team members ,

QMSum_90

false

Yep.

QMSum_90

false

Yep.

QMSum_90

false

and then try to come to decisions uh about the concepts uh you have presented. So and that uh will uh

QMSum_90

false

Mm-hmm.

QMSum_90

false

we have some uh forty minutes uh to complete this uh.

QMSum_90

false

Mm-hmm.

QMSum_90

false

So um who has the fir do you ha Anna do you have your presentation ready ?

QMSum_90

false

I have a presentation , I'm just making this

QMSum_90

false

Yeah I think yeah the Matthew it is it's important that Matthew yeah is here

QMSum_90

false

Okay. Ah

QMSum_90

false

because it's really a a team uh project with a team

QMSum_90

false

there is Matthew.

QMSum_90

false

and if someone is not here then we cannot

QMSum_90

false

Sorry.

QMSum_90

false

but it's okay it's good.

QMSum_90

false

Okay I'll just email you this file , my presentation.

QMSum_90

false

So. Good. Do presentation ready ?

QMSum_90

false

Mm-hmm I'm just emailing it to you.

QMSum_90

false

Oh okay.

QMSum_90

false

So

QMSum_90