0:00:15hello everyone and as you know i'm from the speech technology it sounds and today
0:00:22i'm going two presents some information about c company
0:00:29well
0:00:32the speech technology its answer was found it in a nineteen ninety
0:00:38and it was funded by the roll call for young scientist
0:00:44and today
0:00:47we
0:00:51the company has been are working on the innovation technologies during twenty four more than
0:00:59twenty five years
0:01:01and today we have the
0:01:04why stuff more than a three hundred and fifty a people who are working on
0:01:12a different here we're on the speech technologists
0:01:18the most of them all the scientists and engineers
0:01:23our company has
0:01:25twenty eight p h d degree
0:01:29the people and here
0:01:33the mean are
0:01:35or well
0:01:40a d r o all about set to yell so
0:01:44well and the about the here we have of our research
0:01:49and zero
0:01:51our
0:01:53solutions for the industry
0:01:55we are working on speech recognition speech since it is always and basically a matrix
0:02:01and speech and that in a six
0:02:04well
0:02:07the rap a range or for our interest and the of course we are very
0:02:13interested in developing a good solutions for the
0:02:19voice back down by an entry
0:02:22a well you can see there
0:02:26many of our customers
0:02:29in our in
0:02:31in the side in the rock try or international companies
0:02:37l so we
0:02:39collaborate with the assembly nowhere stays for example
0:02:45well no well no one it's more university we has
0:02:50our department there and
0:02:54we
0:02:57ditch our students and say
0:03:00p h is you dance
0:03:03well
0:03:05also or
0:03:07i
0:03:08as i and all of the you know in the right this cocoa project started
0:03:13a near a few years ago
0:03:16and we participated the up to
0:03:21well this to see manage to build just well irene tedium a in the industry
0:03:29it is the one of the leading parent it in the wall
0:03:32or was then the one hundred and twenty a technical specialists they are
0:03:38i have mentioned it
0:03:40well about the minimum energy management standards
0:03:45we are fully certified
0:03:47i so well nine
0:03:50ninety or one company
0:03:54liz i would it'd
0:03:56financials and the
0:03:58i can say that we have well balanced mentioned to him
0:04:03and
0:04:05this it's
0:04:07about the company just in fewer
0:04:12tension
0:04:20okay given okay
0:04:28thank you mean
0:04:29so now i don't know cuddles
0:04:32it is he's going to present something about zero one week
0:04:44so with modeling every my thank you for coming so i'm gonna talk a little
0:04:48bit about
0:04:50zero is like by and mainly at i think the main purpose of these of
0:04:53this session is to do hear from you so
0:04:57so i had this open question is what that it's what to the role of
0:05:02in this to be in this advancing the indy scientific a community so that's right
0:05:09my performance by
0:05:12trying to
0:05:14set
0:05:15so about zeros like
0:05:18so a lot of
0:05:20maybe some of you know this is a you know supplier of audio and volumes
0:05:27and that was and there it should wait and somewhere solutions so we'd about habitat
0:05:31somewhere
0:05:33in terms of what we morgan a whole chain for all you change from the
0:05:38microphone from these and microphone say at combat pairs dsp to the process of the
0:05:45application process i'm back to the lower speaker and we also what software so it's
0:05:51is
0:05:53it is interesting for us which is
0:05:55no need for signal processing not suppress the on and what in my profession and
0:06:00of course
0:06:01speaker recognition
0:06:03and we focus some of my solo c rs optimal by the time and a
0:06:07can summarily applications and in this feels in speaker recognition is mostly needed for authentication
0:06:13so we you know already we don't focus on politics it's
0:06:18okay or a feel is mostly authentication it's
0:06:22we domingo size components from one house and place it is
0:06:27all spread around the world and us a euro s
0:06:31and it is one thing i want to mention which is i think it's good
0:06:34for this is an american
0:06:38so the court or is very important so we have this kind of slow as
0:06:42like silver from san the you know like that in your when you're right that
0:06:45and you spend a boss dataset and all that stuff so it's just say okay
0:06:51that's then school and no
0:06:54i think within about this is that they are really the really get about the
0:06:57employee so they and this everything for as a selling receptionist because
0:07:02i mean
0:07:03divided we motivated that do in your
0:07:06you or your so if you
0:07:09if you need to be in the scientific community in you need to collaborate with
0:07:13other teams and you need to you know us error and the defence we do
0:07:16here they understand that and they say okay that's it that's the way you work
0:07:20in this way so they don't aids i mean it's
0:07:23it's a matter of you know keeping the ideas of the of the
0:07:29the researchers have otherwise in one
0:07:31the one of the
0:07:33let's see how we set thing in and that's important for this from a combination
0:07:39where cultural is very important
0:07:41so
0:07:43that's all i wanted to say
0:07:44so that all zeros lower one or several power
0:07:48presentation is about the remote of interest in the scientific community so how eight
0:07:54what's are you and i want to select we you and maybe
0:07:57get some feedback any directories
0:08:01let us know how
0:08:03come we had but can we do so
0:08:06so in our previous update in the industry have assessed themes and processed things
0:08:11i idea is that they should have one foot on the on the scientific community
0:08:15and of course of emitting state of the company and in the end to combine
0:08:19is the one can you
0:08:22this of what you're right
0:08:26but egg
0:08:27the but it's we have a so in the industry and let's say that we
0:08:33know
0:08:34that we are more aware of the challenges of the we use cases that a
0:08:43magi day the reset so actually data you select two sources of innovation one is
0:08:48the and what we call technology both and you're one is my
0:08:53so technology post what we generate here so
0:08:56if you're serving bs and we get this is this new technology it's gonna change
0:09:01speaker might be it's gonna be better than before but is not really oriented to
0:09:06any use case or any porpoise but we have also the other which is the
0:09:09market was so maybe it's important to face noise phase reverberation i don't know what's
0:09:15but in s use case what's gonna be
0:09:18how is gonna be the morning five year so that's
0:09:20that's the thing that companies may be may know more about that because they do
0:09:25some complex and they have a marketing change from the inside looking to that so
0:09:29this kind of information
0:09:31and a help us define the challenge s and i think that's that maybe or
0:09:36are our best contribution yes help to define it telling us how things to be
0:09:43evaluate how and
0:09:47once what have a customer a suspect team while a how the future is gonna
0:09:52be how do we see some to set and maybe a
0:09:56contribute to that so that's two main points and
0:10:00something here so
0:10:01so in this case of hadley finding the time yes
0:10:04but also we have a full in the scientific communities i think
0:10:10it's important then we keep it very competitive correlation so if we define very long
0:10:15but long term goals a
0:10:18it several okay framework to collaborate and seventy s a
0:10:22are not very close to the problem so does
0:10:27so that's
0:10:28our bill
0:10:29i want to know
0:10:30so that's what
0:10:31so the role of the industry being the scientific community
0:10:35okay that's all i want to see
0:10:43as follows now it's time and then you will have a few minutes
0:10:47i know you can do it
0:11:02okay
0:11:04i'm gonna be very brief what'll people the top because i think that was your
0:11:07you know idea that we just
0:11:09talking twenty minutes but unfortunately the forensic people be than the low us to detect
0:11:13so much time but
0:11:15so the main the main goal of my speech now it's gonna be just give
0:11:19you a brief introduction recognition because i don't know how many people here is aware
0:11:24of we are we are they and you see we are like the tiny little
0:11:27company between these two other guys we rst like
0:11:31i like this evening like c as like it is small power of x techniques
0:11:35you know that we are resisting to
0:11:37to be guys
0:11:38so like means a sanitised company that was created in two thousand and four and
0:11:43it was a as an a startup of id sc-ubm grouping method
0:11:48in the university and buddy technical university matrix actually it's funny "'cause" now twelve years
0:11:54ago we where publicly presented in a novel use a congress the last one that
0:11:59was organized here is pain and still to this one about back and of an
0:12:03important little
0:12:05so
0:12:06so it's been a long time now we have we have three main locations married
0:12:12of course where the heck what is are we have and of using in washington
0:12:16where we're supporting the us business or whether and capetown where or
0:12:21so that we can guys are helping us to solve problems like you collaborate and
0:12:26so on
0:12:27in total we are around twenty five people
0:12:31as on a small company we are still growing and in fact that we lost
0:12:37some people not so long ago move into two zero is because we decided to
0:12:43just so sailor or
0:12:47they mobile business of the company to zeros and now all these guys are
0:12:53are working with them
0:12:55and we are probably the most doesn't think that agnitio has is that we are
0:13:00probably the only company out there that is fall fully focus on automatic speaker recognition
0:13:05that's all we do we don't do a speech
0:13:08we don't do any other language-related a technology we will be bored i in both
0:13:15paramedics so we train would try to be as good as good as we can
0:13:19for that
0:13:21and today a or b guest book pretty nice is still in the in the
0:13:25global market because of it with born we porting the forensic community because actually the
0:13:30very large enough the company's a joint program between the ubm the university madrid and
0:13:36what if you need in this pain that they just a to invest some money
0:13:39in the group and so on and that's
0:13:41the a priori enough of the goal but actually our forensic two
0:13:45it was the very first problem we developed for them
0:13:49actually i do using previously
0:13:52that they the people talk about the level for quite awhile
0:13:58during the presentations today according to the last interpol survey that the backend lightyear
0:14:07we are around roughly fifty percent of the for in c class that are doing
0:14:13the automatic speaker recognition out there
0:14:15so
0:14:16in the case then we are finally we
0:14:18we got some traction and some
0:14:21some success can be seen people that we are good
0:14:24and i case and i don't want to extend much more you are interesting the
0:14:30problem we're gonna be layer in the
0:14:32in the silent so until the session so more than happy to show you what
0:14:37we control because we have problems about also institutional but for those that
0:14:41that we can we will
0:14:43and
0:14:44and as and in fact
0:14:47we are
0:14:49rate full of well being here
0:14:51the embedded because
0:14:52actually our view it's pretty close to what kind of the explain that we are
0:14:58truly believers of the these become data give a community and we try to share
0:15:05as much as we can then we try to the us open as
0:15:08as we can because we really think that is the only way to make these
0:15:11technology a evolve and the good thing is that we can provide
0:15:16some insights from reality that sometimes is important because sometimes you guys need some ground
0:15:22to two types and to understand what are the real problem so there and not
0:15:25only just reduce that point one in the equal error rate
0:15:30because maybe you are you're just reducing point one and equal error rate and you
0:15:34are multiplied by ten the memory that you're using so maybe that's important also to
0:15:38have that that'll to be that is okay and it works great but i wanted
0:15:42to work in a laptop or
0:15:44or like these guys of y two working now mobile phone that is even more
0:15:48difficult you know
0:15:50so
0:15:51so and that's why we keep we keep a very close relation with many with
0:15:56many people actually around and
0:15:59and
0:16:00and that's pretty much it and i hope that
0:16:05you tell us because i want this to be not for some very long but
0:16:10i wanted to be here we meet interactive then we would love to understand what
0:16:14do you guys the spectrum as also because the of course we get advantage of
0:16:19you working on these speaker we learn and then that's also always useful
0:16:24but we also want to learn from you know and that's why i think be
0:16:28that would opportunity to be here and discuss
0:16:31thank you
0:16:38that sounds and i think that basically the view that follows the is more or
0:16:44less
0:16:45a lot of components are
0:16:47unless he responded so basically i think that we have to the nominees
0:16:52to listen you and
0:16:54or to listen that cutting
0:16:56so the first question or the first topic we can we can talk about not
0:17:01a long time
0:17:02e c and
0:17:04how in the how i can only see the relation between economy and just and
0:17:09then we can discuss some of that if you used in these interesting
0:17:13so
0:17:15anyone wants to talk
0:17:22okay
0:17:24for lid
0:17:32yes it is easy to explain what academic a expecting from you it's morning
0:17:38so you know that but
0:17:41to be a little bit more precise we need more may to do open research
0:17:46we have good relation was industrie for applied research we all missing all of this
0:17:52morning for
0:17:55new innovation you know ten years we have movies unimodal because whole one of the
0:18:01morning we have is that the two
0:18:04close them on a project so if you want to help of you academy just
0:18:09you to give or information b s p l c formation was maybe no more
0:18:15than four pitch you can be your
0:18:17and just if we if we to use the remote this he asked to for
0:18:22a competition we would propose
0:18:25the topic and you will to give a paid for the kids t
0:18:30in this case we could have really open topics
0:18:34and the we could make the market for you in ten or twenty use but
0:18:39so
0:18:44okay any of you want to
0:18:46the player to just comment
0:18:53okay that's
0:18:55that's a little bit
0:18:57that money is always select the tickets over some four
0:18:59compensate i mean they are they are therefore probably we don't forget that so
0:19:07i we can do that we can do that but it's the subways i mean
0:19:14when you
0:19:15i mean when accompanist a what you money somewhere
0:19:19he they expect some
0:19:21some redone so we have some collaboration with big universities a
0:19:26going on which is really based on subcontract in and
0:19:31leading then and go ahead go ahead with research and so on
0:19:35but it's
0:19:36it's actually unless it could be it could be possible that you
0:19:41several companies get together and agreed to find something like
0:19:44but you're mentioning its i don't see why couldn't be done i agree with that
0:19:50but if there is only a few of then it's
0:19:53it's a little bit controversial because you're funding for the comparison as well so that's
0:20:01that's always complicated doesn't obligated is so components is only one
0:20:07a star finally in some projects with for the academia but there are
0:20:12involve a small groups maybe some
0:20:15not is not quite open
0:20:17but it could be the case that we could get together and five something bigger
0:20:22not would be and i'm not directly but
0:20:25they are real that
0:20:30and i would like to add a due to that
0:20:34that
0:20:36strongly i strongly agree with you at that we need to give you long term
0:20:41because i think that is t and i think that these myself i do and
0:20:45i pretty for the disguise also have pretty clear where we would like to go
0:20:49and ten years
0:20:52is true that
0:20:53that as a scholar said we need to find a balance between them between investing
0:20:58in research open investigation and the party that any company wants to wants to want
0:21:06to get right so that but i think that i and actually i think that
0:21:10different companies over here
0:21:12we are kind of all band and we try to parties actually which we try
0:21:15to but was as much as possible that sometimes believe me is not so easy
0:21:18now company just to get the also relation just to get popping whatever you are
0:21:23getting right so
0:21:24but i think it's all about the find a balance because the at some point
0:21:28we want to help and then in the other and the other hand
0:21:33the owners of our companies want to get some profit for from that right so
0:21:37that's a key i think that find that balance between these two things
0:21:46i would i would like to say well of course have to agree completely with
0:21:51some possible to apply a dream for us but anyway a there's another
0:21:58option that day starts to be almost mandatory it want to get funding from the
0:22:03european union of these that most more involvement
0:22:06on the core of the companies only thing
0:22:10initiatives
0:22:10because with a i'm talking about you from a europe is pretty perspective
0:22:16with f p seven framework programme seven of the european union
0:22:20a the big projects where mostly the well it rate it will escort when huge
0:22:27enrollment all companies was present with a rice and twenty the new paradigm for your
0:22:34parents funding
0:22:35the involvement of industry have become has that's even increased
0:22:39and the many of the proposals have rejected because they are not space that the
0:22:45that those places this a but they are asking for companies to lead the proposed
0:22:49accordingly the proposals to get more involvement in the proposal
0:22:53not just as integrators or facilitators but just four
0:22:58doing the research themselves
0:23:01and generate a products that are
0:23:04i the they're going to be integrated and on a high-level at entertainment as it
0:23:09a little bit
0:23:10so
0:23:11perhaps a good the mum
0:23:15middle point between something or research
0:23:19totally and a doing what we're doing currently which is collaborating as a subcontract would
0:23:25be to get the mobile then in this kind of initiatives with which of course
0:23:28there risque because sometimes they are not found that many times the other funny there's
0:23:32truly competitive
0:23:34at least in my i'm at the environment around
0:23:37most of the time i see a big project grant it is because the companies
0:23:43cody at eight we have this speaker identification integrated project now from f p seven
0:23:48that is correlated by parent
0:23:51that's a good example that's when companies gets a huge involvement of this then we
0:23:56can expect to have a lot the funding for lots appear the students and whatever
0:24:00else but it eight in it it's a for in my opinion it
0:24:03it makes sense that the companies assume that risk that i would like to hear
0:24:08the pin the of the companies
0:24:11yes then you have to code in the template so induce the academy and government
0:24:16or european okay
0:24:19so many common about how government or you'll open that case
0:24:24so that's
0:24:25i mean i fully agree i in fact
0:24:29assets the company recognition is a
0:24:31we think that we are we're getting to h twenty that we didn't we didn't
0:24:34do enough if be sure that and i'm gonna talk now from a comedy or
0:24:39perspective not from being just in general because i'm i don't know many others but
0:24:44for is huge afford only because as a twenty five people company and
0:24:51be part of the combination of a project like these that you know how we
0:24:55these that usually both like thirteen different companies from seven different countries you know that
0:25:01it's creating only that discrete in like point emails per day and
0:25:06and let a mum is it's tough stuff but this is so that we want
0:25:11to be there and we're moving in that direction but is to
0:25:16i don't know if this so much about
0:25:19have been a common a big corporation the leaving the a the initiative it's more
0:25:24about
0:25:25been able to percent rio the usability for whatever you but we i think that
0:25:31that's exactly that i think that the european union is able eighteen very possibly his
0:25:36that
0:25:36the capability that you have
0:25:38to put whatever you the below
0:25:40i'll they're doing things
0:25:43use useful things right and i think i think that of course for that and
0:25:47industry should be should be also uses there are some in users also or here
0:25:53and i think that it's also important that they get involved as well in the
0:25:58case of for in six the forensic labs in the case of
0:26:00some other and billions the corpus is the government agency in the case they of
0:26:09i identification of all of citizens the proper government brian so i think it's also
0:26:17three party thing is not only about the you guys researchers we guys the company
0:26:23but also the in users that also be about the about proposed to the project
0:26:31i am also
0:26:33okay like to comment it's
0:26:36i completely agree with you but it there's
0:26:39one problem we also hired in this in discourse a it's
0:26:43that is that it's a lot of overhead
0:26:46yes for going to the court so that's
0:26:48that's a big running takes a long time a needs
0:26:52sometimes from the company
0:26:53point of view c then someplace if it's even easier to identify
0:26:57the rather than
0:26:59you know that go ahead then try to spend all these maybe from nothing
0:27:03the problem the problem you
0:27:05you have they're also is
0:27:08is that
0:27:10and i think essential we have in a in our side in the industry here
0:27:13we have to fix it which is a
0:27:16we should look into this pre-computed discolouration more but in the industry side so i
0:27:22i always see this okay this competition in this utopia riots where you have one
0:27:28or two company leading someone prior than the other two competing and it's maybe for
0:27:32this one or for the other one and they don't get together and they have
0:27:35a different same thing so it's
0:27:38so that something some we may need to fix on our side so far were
0:27:42computed experiments but i
0:27:43i four that we would you that that's
0:27:45problem either graph to get
0:27:47to get well all we want
0:27:50we all
0:27:55okay any question
0:27:57or comments down
0:28:04so i ask a question i'm not sure if you're grew up there probably you
0:28:08be unhappy with his question but
0:28:10if you're looking at maybe in the feeling forensics and so forth there are researchers
0:28:16the mean i in this room especially in speaker recognition space that in focus more
0:28:22on the sre
0:28:24but i think if you look at bs very in you look at maybe some
0:28:27of the work in the forensic space there there's some more at their connected but
0:28:31disconnected as well and also think that the ability to get access to realistic data
0:28:37in the field is very challenging i mean it's not easy to get her case
0:28:42data
0:28:43to be publically released so i'm wondering from a company perspective
0:28:48the willingness of your organisations to be able to share your data even if it
0:28:53means it's in a controlled setting maybe of some type of work
0:28:57other user license that allows individual labs to get access to that i would be
0:29:03very important because i think as dark pointed out you know data drives a lot
0:29:09of our work and i still think that there is a somewhat of a voice
0:29:13separation if you will between the technologies
0:29:16and the forensic speech scientist and i think access to the data will actually help
0:29:20people on the technology side released on the research side in a lot better way
0:29:29we're going
0:29:32so
0:29:33a probably this is the politically incorrect to say here but i fully agree that
0:29:38most of the research community for the last five years
0:29:43in my opinion is to
0:29:46sre by s
0:29:48because the most of the algorithm that you find here are focus on optimize whatever
0:29:54we sold you get in the whatever condition of whatever evaluation all mixed and
0:30:00and is true that
0:30:02reality differs a lot from that i mean it's a good
0:30:06you try having everybody log data and anything that needs has been doing a huge
0:30:12work to provide data to everybody
0:30:14but of course that is only representing a small portion of the different conditions that
0:30:19you can find out there
0:30:20and even for as we struggle to get data like that and of course we
0:30:26have our own our own databases we try to build them but
0:30:30but sometimes
0:30:31is not chasing because most of the in users and for the morning for in
0:30:35six
0:30:36they are not so easy to get that data to share everybody and for s
0:30:41and it's true that they are like a few initiatives from the use of for
0:30:46instance what inferior the created a database to really make a i think that they
0:30:51the d the something similar
0:30:54but is to that
0:30:55the idea still it's like what in jail was present in it it's just elected
0:31:00a small databases is more for and i know and you know that there is
0:31:04tong update after real data out there sitting
0:31:08sitting down and
0:31:10and is true that we need to do something because it is true that the
0:31:13end users
0:31:15that's not help
0:31:17both community i mean you guys m and as guys it's very high it will
0:31:21happen to advance and we can get
0:31:25perfect in the in condition a b and c all means there's hardly twenty overall
0:31:31weight that
0:31:32what that means for really make sure so i know obviously if you were between
0:31:37us or eer we see the movement if you're working at work or
0:31:42mated looked at for speech no wild infrared dark speech to all receive are moving
0:31:47things into more naturalistic data but you believing those corpora they show variability but they
0:31:54don't be me not necessarily reflect word you would see sometime german in the voice
0:32:00forensic side of things ensures
0:32:02somehow would be nice you can see more about him and i'll just full disclosure
0:32:07unbiased because with the ozarks speaker committee responsible for trying to work to find out
0:32:13word data could be used for the or sex or
0:32:22resisting
0:32:24you know in our company would be always the something agreement with the customers and
0:32:28we
0:32:29and not i don't know abilities to redistribute usually this to you builds is data
0:32:36for the
0:32:38and community but
0:32:40i agree with the u and of course i would like to sing these survive
0:32:45for the speaker went in the wild challenge because it's which was agreed to punch
0:32:50unity to investigate this key some data just noisy and or different conditions
0:32:59in case of our company we also participate in the air but
0:33:05the data we have is the usually unlabeled to and what we have also have
0:33:12some issues how to use these in our research number topics and we usually
0:33:21can get only the models will we can get the
0:33:27we forms all the data so and
0:33:33in my opinion it is maybe it's a good idea to share i-vectors why not
0:33:41because no huge just a minute it's possible
0:33:46well
0:33:49okay so
0:33:51we all wear what we'll over a decision is
0:33:55for surplus have you hear
0:33:57and thank you your opinion and comments