i will show the also has a three minute warning okay uh i can just and the fifty minutes presentations not tick oh it's about exist display yeah yeah hi everybody have time to to be yeah um uh for the morning not to use a scroll do but thank you so speak to yeah function that says just as shown so a the mother used mission um will yeah yeah that's use a level of used uh it was not some a question that it was a a a a a a the question of a review of a of uh this paper and i even if may looks like a bunch question right question i is is the worst uh use one H or so i we just take the first minutes you know to to give you some context and try to send are uh answer this question so the voice production is mainly made of the three components of the glottal source made uh generated of the larynx level of the uh the vocal tract filter are on finite the radiation of the most level and then leading of the vocal a use a propagated with form i i can be perceived from someone i my feel of research is the voice transmission thus the question is or to transform a probably with form you know that to make it perceived differently in of voice qualities in terms of timber a in terms of uh any other uh perceived elements of the voice and the one uh uh i D is you split of the propagated located where for in order to retrieve a uh the elements of the fun and of an elements of the voice uh and in any other the all sort of a bit up to some the addition but i we use it in order to do this uh yeah uh it is necessary to in but as the voice production which is what a tricky problem and one solution if we need to invert uh uh a search problem uh is to add some constraints and uh for of this and the and models are very useful because the that described an i to go a waveform on add some constrained for these and so the question is now how to estimate the shape by does or for both of model so first i will briefly describe the model or the rows production are used as i say that there is a but also has of a trip fit on addition duration the but source is made of and a shape in the time domain this should as obviously a time position and then we assume that is shape is periodic then uh we assume that the vocal tract is minimum phase or the zeros of these that from it's is is that transform uh lies in side the unit circle and a fun i you we assume that the relation can be simply my of the using a a a time a simple time to do the derivative of which is in frequency domain uh G W i have had the find and then we use a a a each uh representation uh it may that's as also the and um a monies is simply uh defined as H so you will find the now a a a a a phase component as as a complex six L also H C and uh a uh it's or addition is simply we use to uh G H so the proposed was in this paper is the function of phase distortion what is that uh yeah I Ds a first it describe you see you know a D question on describing the do the function of phase distortion and uh of the first I used to remove and a medium phase components or from the main argument X H using a a a a a a a it's uh um me "'em" phrase weight values editions through the real cepstrum then right this um this way of for yeah removing a new face uh real ready helps for example uh uh to remove the contribution in the face uh of the vocal tract fit uh which could be present in the argument X H then we use the seven or the difference about brought or in order to remove and we now phase component remaining in a uh this and division and finally we use a and the two difference a parameter you know to obtain a a a similar addition to the group do we each is a a a a a meaningful in terms of phase distortion so that to the uh example or of those for a function of discussion maybe you at about uh thirty G from but that model which uh described uh the shape of of the a a a a a lot of buttons and we can use uh simple powerpoint use of this but the model a using the transformed the lead you can from a model and this problem to the of the which is used is a a a all the it part of the is more a uh of the pose is uh looks more like any buttons and he this bomb to is beak it more looks like a sign is i a but the right a about on of uh was slide you would see a um a plot which describe a function of phase distortion for are on of the first three harmonics uh with respect to uh its shape parameter uh so this was function or a a uh in that i don't or a many uh elements and there are only related to the shape the i the and on the time position times to this end and all phase difference the and the but i'm to the amplitude things as an an addition by you mean "'em" phase very efficient i and the on to the duration of the buttons seems uh uh we uh walk on a uh a and money model and find the are independent or the new face call component so those function and the you related to as the uh a shape of the buttons and application is i those function of a distortion is the estimation of a of V R D problem at also a jack can from model and in a lot to do this i i will be describe a uh they're could an which is used uh for of these uh estimation which is the great in a phase musician which as been a really use a a you know that to the done line uh instant of see if you can X station or first we use the convolutive residual a which is is the division spectral domain or or was the observed spectrum by model then a uh we can say that if the model correspond to zap some signal it can that the comet quality what do that a is equal to one but that's i can see is which you mean that it's amplitude spectrum is equal to one a and needs space spectrum is equal to zero so the idea of great and phase musician is first to ensure that the amplitude spectrum of the convolutive residual is constant and then a the simple idea is to minimize a face but some you know the to feet the parameter or to the observe scene so we can see that uh a given a of the voice production model as described before no uh the linear phase but the shape of the glottal pulse the vocal tract it under addition and a the function of phase discussion we propose we can see that as functions can be applied to a a a a the convert you raise your are we know scrap the the question but you can just see of the file and uh for the last line but that you can see a representation of the function of its station by means of C yeah division by the you and phase realisation of the numerator or with a an additional in a phrase and this no phase is we be removed thanks to this this and uh all the phase different such as a we can um minimize and the phase spectrum or of the convolutive residual by minimizing is this um L function which is done and the speed is square for mean squared phase using in so one the phase difference which is simply the square of uh a function of phase distortion of the types of seeing our by uh the lf model yeah is some example of and those uh all functions we uh three synthetic signal i'm as you can see uh those functions are uh quite uh uh simple uh uh to optimize in terms of finding being the row the global uh a minimum and describe a minimum uh seems to to close spend while to two of these syntactic video which is used in this example there's we use a simple a um brands method you know that to retrieve the block but the global minimum of those functions in terms of evolution of these uh make that's first to a simple example the well i'm about represent a and requiring a problem of of rose going from a right about right voice to a tense voice um in a you can see and um the yeah ad you and or which is a predicted from the Z to the graphic see no and uh a well you can see uh the a deep part two which is estimated through you you D square based made obviously there is a big yeah has even if at the shape of a use uh estimates modes are quite similar there is a big B be could the link the link between uh the work of for motion the make an mechanical of motions and uh the battle per season not is not to use so we can only observe so colour relation and below low you can see in figure or a a a a a small segment of speech with writing uh estimate of our department um more than a few examples here is an evaluation um made from uh data basis using a game addict or blood the graphic signal for the open question ease of predicted from the uh are D shape part of the and then the idea of a open question is predicted from the click or of got sick and and the figure out just below show we use a standard deviation of these open cushion you know for for different that the base and a as you can see a the proposed was method seems to have a on the a state of the main state of the also made a so yeah just as a mean to our conclusion we propose a function of phase distortion which are really did and mean to the shape of the problem birds and as just function can be used so to estimate a a should parameters of local models thank thank for the action thank you and to yes a lot of time for question i start the discussion with you in it used uh you model by decomposing uh the total transfer function the product of the glottal model the vocal tract and radiation model and you mate the standard assumption about the vocal tract and radiation in terms of as of their face function being minimum phase so be just the tree a make a term so that's that assumption faq you of further analysis at all because you set funds you independent of the minimum of component so uh what if there are to track not but not be minimum face in the end yes obviously use these and make use of a track is fine not mean will phase it will uh a lot in found to the the result of uh the of the the estimate of the of the shape parameters and that the same at a are some shot about the vocal tract filter a is quite is not that a a strong and it seemed that uh uh sit assuming that is me more phase is quite uh a a a a a a good assumption but uh uh i i had made that's for the addition of the mouse a level uh uh it will been may be necessary to make some more about improving the model of the of the models are addition and put that explain the differences you see between us to mets send the class uh cut the graphic estimate yeah yes and i didn't because um you have to hmmm and to remind that uh the egg that a click sorry electroglottographic signals are only linked to the motion of the vocal folds well what we are trying to retrieve a is an estimate of uh an approximation of the glottal flow we and and the acoustic be could between the two is uh for form of views not only now and so we can not only observe a correlation so that's why there is a so much difference yes piece because there thanks and that would be your evaluation you compare against the i i i i a i have method this you don't actually you don't say boss privatisation that's of was used to extract a Q which i think might be chris go and can you can use a but last method used after using D i i have that it's just this is a a a a a uh i i i if is first to get it to estimate the plot of for on not to estimate part of more than sure so uh the I is for to estimate the of for and then to fate a a a a a a a a model on some pulses uh a on the estimated uh a lot of rule using a a uh gradient descent or any you know those uh made we can use just just as a from from my screen's the at that the method jeez "'cause" it it it it the make the method use can to different methods use very different a key value so have yeah just "'kay" thank you you have more questions this is not the case let's say thank the speaker again