elplaguister/DTS_line_datasets · Datasets at Hugging Face

topic_shift bool 2 classes	utterance stringlengths 1 7.9k	session_id stringlengths 7 14
false	Yeah.	QMSum_86
false	Um. So Uh I just sorta think we need to explore the space. Just take a look at it a little bit.	QMSum_86
false	Mm - hmm.	QMSum_86
false	And we we we may just find that that we 're way off.	QMSum_86
false	OK. Mm - hmm.	QMSum_86
false	Maybe we 're not. You know ? As for these other things , it may turn out that , uh , it 's kind of reasonable. But then I mean , Andreas gave a very reasonable response , and he 's probably not gonna be the only one who 's gonna say this in the future of , you know , people people within this tight - knit community who are doing this evaluation are accepting , uh , more or less , that these are the rules. But , people outside of it who look in at the broader picture are certainly gonna say " Well , wait a minute. You 're doing all this standing on your head , uh , on the front - end ,	QMSum_86
false	Yeah.	QMSum_86
false	when all you could do is just adjust this in the back - end with one s one knob. "	QMSum_86
false	Mm - hmm.	QMSum_86
false	And so we have to at least , I think , determine that that 's not true , which would be OK , or determine that it is true , in which case we want to adjust that and then continue with with what we 're doing. And as you say as you point out finding ways to then compensate for that in the front - end also then becomes a priority for this particular test ,	QMSum_86
false	Right.	QMSum_86
false	and saying you don't have to do that.	QMSum_86
false	Mm - hmm.	QMSum_86
true	So. OK. So , uh What 's new with you ?	QMSum_86
false	Uh. So there 's nothing new. Um.	QMSum_86
false	Uh , what 's old with you that 's developed ?	QMSum_86
false	I 'm sorry ?	QMSum_86
false	You OK. What 's old with you that has developed over the last week or two ?	QMSum_86
false	Mmm. Well , so we 've been mainly working on the report and and Yeah.	QMSum_86
false	Mainly working on what ?	QMSum_86
false	On the report of the work that was already done.	QMSum_86
false	Oh.	QMSum_86
false	Um. Mm - hmm. That 's all.	QMSum_86
false	How about that ? Any - anything new on the thing that , uh , you were working on with the , uh ?	QMSum_86
false	I don't have results yet.	QMSum_86
false	No results ? Yeah.	QMSum_86
false	What was that ?	QMSum_86
false	The the , uh ,	QMSum_86
false	Voicing thing.	QMSum_86
false	voicing detector.	QMSum_86
false	I mean , what what 's what 's going on now ? What are you doing ?	QMSum_86
false	Uh , to try to found , nnn , robust feature for detect between voice and unvoice. And we w we try to use the variance of the es difference between the FFT spectrum and mel filter bank spectrum.	QMSum_86
false	Yeah.	QMSum_86
false	Uh , also the another parameter is relates with the auto - correlation function.	QMSum_86
false	Uh - huh.	QMSum_86
false	R - ze energy and the variance a also of the auto - correlation function.	QMSum_86
false	Uh - huh. So , that 's Yeah. That 's what you were describing , I guess , a week or two ago.	QMSum_86
false	Yeah. But we don't have res we don't have result of the AURO for Aurora yet.	QMSum_86
false	So.	QMSum_86
false	We need to train the neural network	QMSum_86
false	Mm - hmm.	QMSum_86
false	and	QMSum_86
false	So you 're training neural networks now ?	QMSum_86
false	No , not yet.	QMSum_86
false	So , what wha wh wha what what 's going on ?	QMSum_86
false	Well , we work in the report , too , because we have a lot of result ,	QMSum_86
false	Uh - huh.	QMSum_86
false	they are very dispersed , and was necessary to to look in all the directory to to to give some more structure.	QMSum_86
false	Yea	QMSum_86
false	So. B So Yeah. I if I can summarize , basically what 's going on is that you 're going over a lot of material that you have generated in furious fashion , f generating many results and doing many experiments and trying to pull it together into some coherent form to be able to see wha see what happens.	QMSum_86
false	Hm - hmm.	QMSum_86
false	Uh , y yeah. Basically we we 've stopped , uh , experimenting ,	QMSum_86
false	Yes ?	QMSum_86
false	I mean. We 're just writing some kind of technical report. And	QMSum_86
false	Is this a report that 's for Aurora ? Or is it just like a tech report for ICSI ,	QMSum_86
false	No.	QMSum_86
false	Yeah.	QMSum_86
false	For ICSI.	QMSum_86
false	or ? Ah. I see.	QMSum_86
false	Yeah.	QMSum_86
false	Just summary of the experiment and the conclusion and something like that.	QMSum_86
false	Yeah.	QMSum_86
false	Mm - hmm.	QMSum_86
false	OK. So , my suggestion , though , is that you you not necessarily finish that. But that you put it all together so that it 's you 've got you 've got a clearer structure to it. You know what things are , you have things documented , you 've looked things up that you needed to look up.	QMSum_86
false	Mm - hmm.	QMSum_86
false	So that , you know so that such a thing can be written. And , um When when when do you leave again ?	QMSum_86
false	Uh , in July. First of July.	QMSum_86
false	First of July ? OK. And that you figure on actually finishing it in in June. Because , you know , you 're gonna have another bunch of results to fit in there anyway.	QMSum_86
false	Mm - hmm.	QMSum_86
false	Mm - hmm.	QMSum_86
false	And right now it 's kind of important that we actually go forward with experiments.	QMSum_86
false	It 's not.	QMSum_86
false	So so , I I think it 's good to pause , and to gather everything together and make sure it 's in good shape , so that other people can get access to it and so that it can go into a report in June. But I think to to really work on on fine - tuning the report n at this point is is probably bad timing , I I think.	QMSum_86
false	Mm - hmm. Yeah. Well , we didn't we just planned to work on it one week on this report , not no more , anyway. Um.	QMSum_86
false	But you ma you may really wanna add other things later anyway	QMSum_86
false	Yeah. Mm - hmm.	QMSum_86
false	because you	QMSum_86
false	Mmm.	QMSum_86
false	There 's more to go ?	QMSum_86
false	Yeah. Well , so I don't know. There are small things that we started to to do. But	QMSum_86
false	Are you discovering anything , uh , that makes you scratch your head as you write this report , like why did we do that , or why didn't we do this ,	QMSum_86
false	Uh.	QMSum_86
false	or ?	QMSum_86
false	Yeah. Yeah. And Actually , there were some tables that were also with partial results. We just noticed that , wh while gathering the result that for some conditions we didn't have everything.	QMSum_86
false	Mmm.	QMSum_86
false	But anyway. Um. Yeah , yeah. We have , yeah , extracted actually the noises from the SpeechDat - Car. And so , we can train neural network with speech and these noises. Um. It 's difficult to say what it will give , because when we look at the Aurora the TI - digits experiments , um , they have these three conditions that have different noises , and apparently this system perform as well on the seen noises on the unseen noises and on the seen noises. But , I think this is something we have to try anyway. So adding the noises from from the SpeechDat - Car. Um.	QMSum_86
false	That 's that 's , uh that 's permitted ?	QMSum_86
false	Uh. Well , OGI does did that. Um. At some point they did that for for the voice activity detector.	QMSum_86
false	Uh , for a v VAD.	QMSum_86
false	Right ? Um.	QMSum_86
false	Could you say it again ? What what exactly did they do ?	QMSum_86
false	They used some parts of the , um , Italian database to train the voice activity detector , I think. It	QMSum_86
false	Yeah. I guess the thing is Yeah. I guess that 's a matter of interpretation. The rules as I understand it , is that in principle the Italian and the Spanish and the English no , Italian and the Finnish and the English ? were development data	QMSum_86
false	Yeah. And Spanish , yeah.	QMSum_86
false	on which you could adjust things. And the and the German and Danish were the evaluation data.	QMSum_86
false	Mm - hmm.	QMSum_86
false	And then when they finally actually evaluated things they used everything.	QMSum_86
false	Yeah. That 's right. Uh	QMSum_86
false	So Uh , and it is true that the performance , uh , on the German was I mean , even though the improvement wasn't so good , the pre the raw performance was really pretty good.	QMSum_86
false	Mm - hmm.	QMSum_86

false

Yeah.

QMSum_86

false

Um. So Uh I just sorta think we need to explore the space. Just take a look at it a little bit.

QMSum_86

false

Mm - hmm.

QMSum_86

false

And we we we may just find that that we 're way off.

QMSum_86

false

OK. Mm - hmm.

QMSum_86

false

Maybe we 're not. You know ? As for these other things , it may turn out that , uh , it 's kind of reasonable. But then I mean , Andreas gave a very reasonable response , and he 's probably not gonna be the only one who 's gonna say this in the future of , you know , people people within this tight - knit community who are doing this evaluation are accepting , uh , more or less , that these are the rules. But , people outside of it who look in at the broader picture are certainly gonna say " Well , wait a minute. You 're doing all this standing on your head , uh , on the front - end ,

QMSum_86

false

Yeah.

QMSum_86

false

when all you could do is just adjust this in the back - end with one s one knob. "

QMSum_86

false

Mm - hmm.

QMSum_86

false

And so we have to at least , I think , determine that that 's not true , which would be OK , or determine that it is true , in which case we want to adjust that and then continue with with what we 're doing. And as you say as you point out finding ways to then compensate for that in the front - end also then becomes a priority for this particular test ,

QMSum_86

false

Right.

QMSum_86

false

and saying you don't have to do that.

QMSum_86

false

Mm - hmm.

QMSum_86

true

So. OK. So , uh What 's new with you ?

QMSum_86

false

Uh. So there 's nothing new. Um.

QMSum_86

false

Uh , what 's old with you that 's developed ?

QMSum_86

false

I 'm sorry ?

QMSum_86

false

You OK. What 's old with you that has developed over the last week or two ?

QMSum_86

false

Mmm. Well , so we 've been mainly working on the report and and Yeah.

QMSum_86

false

Mainly working on what ?

QMSum_86

false

On the report of the work that was already done.

QMSum_86

false

Oh.

QMSum_86

false

Um. Mm - hmm. That 's all.

QMSum_86

false

How about that ? Any - anything new on the thing that , uh , you were working on with the , uh ?

QMSum_86

false

I don't have results yet.

QMSum_86

false

No results ? Yeah.

QMSum_86

false

What was that ?

QMSum_86

false

The the , uh ,

QMSum_86

false

Voicing thing.

QMSum_86

false

voicing detector.

QMSum_86

false

I mean , what what 's what 's going on now ? What are you doing ?

QMSum_86

false

Uh , to try to found , nnn , robust feature for detect between voice and unvoice. And we w we try to use the variance of the es difference between the FFT spectrum and mel filter bank spectrum.

QMSum_86

false

Yeah.

QMSum_86

false

Uh , also the another parameter is relates with the auto - correlation function.

QMSum_86

false

Uh - huh.

QMSum_86

false

R - ze energy and the variance a also of the auto - correlation function.

QMSum_86

false

Uh - huh. So , that 's Yeah. That 's what you were describing , I guess , a week or two ago.

QMSum_86

false

Yeah. But we don't have res we don't have result of the AURO for Aurora yet.

QMSum_86

false

So.

QMSum_86

false

We need to train the neural network

QMSum_86

false

Mm - hmm.

QMSum_86

false

and

QMSum_86

false

So you 're training neural networks now ?

QMSum_86

false

No , not yet.

QMSum_86

false

So , what wha wh wha what what 's going on ?

QMSum_86

false

Well , we work in the report , too , because we have a lot of result ,

QMSum_86

false

Uh - huh.

QMSum_86

false

they are very dispersed , and was necessary to to look in all the directory to to to give some more structure.

QMSum_86

false

Yea

QMSum_86

false

So. B So Yeah. I if I can summarize , basically what 's going on is that you 're going over a lot of material that you have generated in furious fashion , f generating many results and doing many experiments and trying to pull it together into some coherent form to be able to see wha see what happens.

QMSum_86

false

Hm - hmm.

QMSum_86

false

Uh , y yeah. Basically we we 've stopped , uh , experimenting ,

QMSum_86

false

Yes ?

QMSum_86

false

I mean. We 're just writing some kind of technical report. And

QMSum_86

false

Is this a report that 's for Aurora ? Or is it just like a tech report for ICSI ,

QMSum_86

false

No.

QMSum_86

false

Yeah.

QMSum_86

false

For ICSI.

QMSum_86

false

or ? Ah. I see.

QMSum_86

false

Yeah.

QMSum_86

false

Just summary of the experiment and the conclusion and something like that.

QMSum_86

false

Yeah.

QMSum_86

false

Mm - hmm.

QMSum_86

false

OK. So , my suggestion , though , is that you you not necessarily finish that. But that you put it all together so that it 's you 've got you 've got a clearer structure to it. You know what things are , you have things documented , you 've looked things up that you needed to look up.

QMSum_86

false

Mm - hmm.

QMSum_86

false

So that , you know so that such a thing can be written. And , um When when when do you leave again ?

QMSum_86

false

Uh , in July. First of July.

QMSum_86

false

First of July ? OK. And that you figure on actually finishing it in in June. Because , you know , you 're gonna have another bunch of results to fit in there anyway.

QMSum_86

false

Mm - hmm.

QMSum_86

false

Mm - hmm.

QMSum_86

false

And right now it 's kind of important that we actually go forward with experiments.

QMSum_86

false

It 's not.

QMSum_86

false

So so , I I think it 's good to pause , and to gather everything together and make sure it 's in good shape , so that other people can get access to it and so that it can go into a report in June. But I think to to really work on on fine - tuning the report n at this point is is probably bad timing , I I think.

QMSum_86

false

Mm - hmm. Yeah. Well , we didn't we just planned to work on it one week on this report , not no more , anyway. Um.

QMSum_86

false

But you ma you may really wanna add other things later anyway

QMSum_86

false

Yeah. Mm - hmm.

QMSum_86

false

because you

QMSum_86

false

Mmm.

QMSum_86

false

There 's more to go ?

QMSum_86

false

Yeah. Well , so I don't know. There are small things that we started to to do. But

QMSum_86

false

Are you discovering anything , uh , that makes you scratch your head as you write this report , like why did we do that , or why didn't we do this ,

QMSum_86

false

Uh.

QMSum_86

false

or ?

QMSum_86

false

Yeah. Yeah. And Actually , there were some tables that were also with partial results. We just noticed that , wh while gathering the result that for some conditions we didn't have everything.

QMSum_86

false

Mmm.

QMSum_86

false

But anyway. Um. Yeah , yeah. We have , yeah , extracted actually the noises from the SpeechDat - Car. And so , we can train neural network with speech and these noises. Um. It 's difficult to say what it will give , because when we look at the Aurora the TI - digits experiments , um , they have these three conditions that have different noises , and apparently this system perform as well on the seen noises on the unseen noises and on the seen noises. But , I think this is something we have to try anyway. So adding the noises from from the SpeechDat - Car. Um.

QMSum_86

false

That 's that 's , uh that 's permitted ?

QMSum_86

false

Uh. Well , OGI does did that. Um. At some point they did that for for the voice activity detector.

QMSum_86

false

Uh , for a v VAD.

QMSum_86

false

Right ? Um.

QMSum_86

false

Could you say it again ? What what exactly did they do ?

QMSum_86

false

They used some parts of the , um , Italian database to train the voice activity detector , I think. It

QMSum_86

false

Yeah. I guess the thing is Yeah. I guess that 's a matter of interpretation. The rules as I understand it , is that in principle the Italian and the Spanish and the English no , Italian and the Finnish and the English ? were development data

QMSum_86

false

Yeah. And Spanish , yeah.

QMSum_86

false

on which you could adjust things. And the and the German and Danish were the evaluation data.

QMSum_86

false

Mm - hmm.

QMSum_86

false

And then when they finally actually evaluated things they used everything.

QMSum_86

false

Yeah. That 's right. Uh

QMSum_86

false

So Uh , and it is true that the performance , uh , on the German was I mean , even though the improvement wasn't so good , the pre the raw performance was really pretty good.

QMSum_86

false

Mm - hmm.

QMSum_86