Player is loading...

Embed

Embed code

Transcriptions

Note: this content has been automatically generated.
00:00:01
uh_huh
00:00:03
um okay then uh thank you for introduction it's a pleasure to be here representing
00:00:09
a titanium uh in the name of sedition when you wouldn't come and then
00:00:16
and as it was uh uh i would just outline first let this presentation is about first winter disheartening
00:00:24
then i had planted that should talk a bit about augmented reality but
00:00:27
uh i think at this point everyone understand quite well the concept
00:00:32
and then i will say how i would tell howard anyone wants to work with augmented reality and has been working with augmented reality
00:00:39
and also uh some problems as some things that they had tractors
00:00:43
regarding tracking navigation for uh uh look for another reality
00:00:47
finally i would just a overview of some of the current limitations that uh are sometimes are not so obvious but that uh
00:00:54
that uh we expect to be a in a in the near future to be addressed on on the augmented reality a hardware
00:01:02
then to start with a are turning has been founded in two thousand eleven it's
00:01:06
a nonprofit foundation now uh how it's activities uh turns around the motion capture
00:01:11
and then most of it is directed to research and design of my decision and also your karate
00:01:16
then currently we want to bring those things together and try to integrate the research
00:01:20
we haven't both enjoy a single let's say over certain design a topic
00:01:25
then we also provide some services like a recording of motion capture recording for movies and
00:01:30
uh and games we also promote some workshops as it was also mentioned that
00:01:38
then here is just a small piece i would just show up part of the demo reel of of of the opinion
00:01:45
uh and then this uh we were great is a system that actually has recently became
00:01:51
a start to uh a bit here a engineer and also in los angeles
00:01:58
then we had a actually a very diverse topics like on vitamin three and the
00:02:06
and you can find a complete video at a u. two channel for her name
00:02:13
and now you talk more about the medical were certain design or uh what has been done on the menu go uh um
00:02:19
viewed by opening then we focus mostly are actually on court appendix a and then we have a they have little but
00:02:26
uh an approach church riddles uh to uh for simulation of
00:02:30
um motion and and some of potential issues of trends
00:02:34
many consist of a capture of a c. t. scans the number
00:02:37
i the capture of motion uh take the buyer performer
00:02:41
and then the reconstruction uh to the reconstruction of models and the some tissues of one's uh mostly actually
00:02:48
and moreover also uh they were construct kinetic reconstruction because joints are
00:02:53
not the perfect also get a don't have a perfect
00:02:56
was like movement uh they also have translation and then we
00:02:59
have to replicate that through the tracking of extra markers
00:03:05
uh you know the correct way and then simulations of a potential for instance uh contact of puns and
00:03:11
and things that can also a a pain and uh uh or uh or mental conditioning in appendix
00:03:18
uh and then turn that into some clinic obligation that can help
00:03:20
a medical doctor or one for a medical what partners true
00:03:25
uh to diagnose and to treat a some problem
00:03:28
or to a plan a circuit surgical intervention
00:03:35
then here are uh make use of some of uh the replacement has been developed that there
00:03:40
like uh on the left it's a simulation of of for the by mechanical
00:03:44
movement of reflection of any plus some uh tension owned a a ligaments
00:03:51
on the centre it's a of also having a it's a tennis
00:03:55
player recording in which a thing with or we try to
00:03:59
determine that that's the stress that maybe cause it on on
00:04:02
the shoulder joint ditches repetitive uh above the head movement
00:04:08
and on the right it's a tool for the uh actually for for a better understanding how that uh i
00:04:13
trying to behave after uh i hate the trying to behave after complete uh at also a publicity
00:04:21
and on this case we want to help to better plan how to use
00:04:24
a joint during that does a protest the process should replace it
00:04:30
and also what is the correct size for for to of being planned for that specification
00:04:39
then about a maturity you have for a i quite a bit and i believe
00:04:44
uh and uh from the computer science point of view it is defined it
00:04:49
as a as supplementing the real informant your computer generated sensory input
00:04:55
then very commonly uh i guess be stretch if that are true uh a
00:04:59
cellphone or or or some other a portable device uh with a camera
00:05:03
and the image analysis and then i also through projection a projection is actually uh i i guess
00:05:10
well uh one major have a scalable augmented reality when you see of of for instance does matter
00:05:16
my projection in public buildings it that that those are very common in love than for instance
00:05:22
and moreover what uh we are focusing now is on the holographic a. r. which is
00:05:26
bringing the uh trying to make that let's say that the word or image
00:05:31
into something that is seen or felt as if it was a tangible object
00:05:35
and for that you have to have a careful uh you you have to get
00:05:39
have a good celebration of five three screen and also have a very uh
00:05:42
efficient tracking of of of position of point of the position and of of empowerment location
00:05:49
and like on this application which is meant for indications so that you
00:05:53
can see different uh ill layers of a of a border structure
00:06:00
and then uh it i guess uh it's not the first element you also hear about today uh uh this is the
00:06:07
grim project we have been working non yeah it has been funded by a lot too was telling me huh
00:06:13
and uh we got the per position is to develop some a a a are close to help on this uh
00:06:19
on the on surgical planning and also on better compression of anatomy uh and also we're time motion uh
00:06:26
then we went for instance you put the the motion tracking system the patient
00:06:30
uh having this cannot fees on a joint of ones and joint
00:06:34
now you think a performing in front of the doctor to see whether this can bring also better
00:06:38
perception of uh his problem and and but they show their vision but that that is required
00:06:45
then then uh that general terms we are interested in in bringing actually everything that i presented about tightening into
00:06:51
and amplify that if augmented reality to make it more usable and and propose kinda complications if that
00:06:57
but or long term and more specific is of using that on the operating room
00:07:02
then for that we we we expected that basically to to
00:07:07
modify what is to be used as a navigation system
00:07:10
uh it instead of relying on on on a fix it screen we we
00:07:15
would only use of the tracking so that we have a very reliable
00:07:17
tracking of the headset and of the ones that uh on this case on
00:07:22
on the show there are two plus the potential purple or publicity
00:07:27
and uh we have uh actually partner uh i met medical partners in the hostels and also
00:07:33
a possibility now of upper corner to to go on what what one of this project
00:07:40
and of course of materiality considering is not something you this is
00:07:45
a project from the university of most the time in geneva
00:07:48
but as you can see this is based on on uh a projection
00:07:52
any sewing interesting because it it it can be churchyard however information
00:07:56
that is contained i don't on the projection is only a two d. and it's basically a map you don't have the that
00:08:02
then they use additional equipment chew chew things information with with
00:08:05
holographic holographic augmented reality you can actually put everything
00:08:09
in place and have a good compression of the deaf and their actual location of of the
00:08:16
of of the problem play something duration these other one from proper is
00:08:21
also very interesting as a way to visualise the doing the the
00:08:26
the interface duration how the the vessels in the league or distribute it
00:08:33
and now you go to some uh uh problems that you have a better set so far and also some limitations that
00:08:38
she should be kind of application we propose the the getting
00:08:43
tracking of uh all this is not suitable for
00:08:46
um well this tracking is made for recreational use it's made
00:08:50
to to play things in a room size it's uh
00:08:54
informant uh usually things that are somewhat for and the disability of uh often hologram can shift
00:09:00
by a few millimetres or one centimetre which is a lot when you are trying to
00:09:04
to make an intervention you if you wanted for surgical use for instance
00:09:09
and then the map on the top right actually shows that we had the variation in space when comparing an optical tracking system
00:09:16
which is the one that you can see a tree markers aunt is a wellness letter use it as sort of
00:09:21
reflective markers for infrared information and then capture by summer came or we can it is uh the the um
00:09:28
the quality of the position of the subject in with a quality that is below the uh i mean it or
00:09:34
from their below millimetre and then when compared to feel less we had a variation
00:09:38
of up to forty centimetres in a range of uh eight meters uh work
00:09:44
which is a quite a lot
00:09:47
and for that we had a two came up uh uh when we do that there is a different problems that we
00:09:53
the centre of projection of the image in their wallace is sticks it in
00:09:57
the device and we for those two markers we actually great uh or
00:10:01
we create an arbitrary cornet system that we have to match to the one i mean the all
00:10:07
the lines so that as seen on the lower right teenage we can have a perfect
00:10:13
or near perfect overlap a off real information we've written information
00:10:17
in this case we have uh forty millimetre marker which is been overlap but by a forty millimetre weirdo is here
00:10:26
then does objecting me move it around in the scene from different places and this
00:10:29
is a also limitation of the getting system it's not make shipwreck dynamic objects
00:10:34
it's not makes you say you are objects are uh i know that recently the integrated just some with some nice
00:10:40
ago libraries but it's to require a lot of processing for that
00:10:47
and now another issue that actually we we face now because we are going to another path is that we have like in c.
00:10:54
and then for that we have been combining actually some a sensor uh
00:10:58
initial sensory input so that we have a lower uh we just
00:11:04
the again as you can see that the movement of the object represented the quite great amount
00:11:08
of latency but by using the the initial sensor for instance and all as we can
00:11:13
actually cancel out is that because we have a much uh it smaller uh like consume
00:11:20
from the inertial sensor
00:11:23
then this is just a and to a simple project at this age is is not the most current one
00:11:29
but then there is to some like into that indignities that but this isn't the visualisation of uh
00:11:35
of uh of keep and uh they more uh that was actually scan it from this person
00:11:41
and then we had the person performing the the the movement and and
00:11:44
been able to visualise his own bones and he's an anatomy
00:11:48
then we expect you to expand that in order to show
00:11:52
a simulations intro for instance potential conflicts and and
00:11:55
and visions of the one that may be affected by by some signals or by some uh repetitive uh
00:12:03
repetitive motion like in sports medicine so that we can have a more um
00:12:09
customise it treatment for that person
00:12:14
and now i just want to highlight a few limitations that uh uh are not always uh
00:12:21
that we not always realise that they are they are if we want a let's say a
00:12:24
a short term solution in intimate reality then maybe the the main one is that uh
00:12:32
normal screen some uh and projections like on the wall as they are made for a a constant i accommodation
00:12:39
and usually a few meters away on the case of they will last for instance they
00:12:43
do not recommend any all in a hologram that is closer to making centimetres
00:12:48
because at this point if you focus on the hologram you see your handler it for
00:12:52
instance if we're close if you focus on your hand you see the hologram blur
00:12:56
and for for the orange bowl which is a surgical intervention disease a
00:13:02
critical actually because either of them that medical doctor we'll see
00:13:06
they're handing focus or the beach overlaying focus but there are actually a lot of research on a like you displays there
00:13:12
are a couple of companies that are ready and also that they are working augment reality with like two displays
00:13:18
and uh i would bet that microsoft if they will go on what we've always they we work with like to displace true
00:13:26
and another current limitation is uh on the case of the all these is that a few the view is actually very narrowed and
00:13:32
you cannot proceed that much if objects are quite far but if they are we think they're reachable volume
00:13:38
this actually becomes a a problem because you cannot see the entire hologram u. i.
00:13:42
if you cannot see it uh as an entire object this may lose the
00:13:46
the the sensation of sort of guinness on it that is based in space
00:13:50
it's more of a of an open object and that we can object
00:13:55
and finally there are parallels to a lot of features like a rate also true for the view and read issue
00:14:01
uh i and uh it's screening displacement and the interpreter distance calibration precise calibration
00:14:08
a centre of projection of a also possess calibration uh but uh i'm confident that in the next few years uh
00:14:16
those will be that the the solutions that are being proposed to integrate putting in some uh headsets for automatic calibration
00:14:25
and then that's it thank you very much
00:14:34
see
00:14:36
integration what's that
00:14:43
ooh thanks for your talk edges additional question to what extent development
00:14:49
the u. d. which are interesting for redevelopment i'll share though
00:14:55
because to what extent it is 'cause the noise addition europe for yourself to what extent as can be
00:15:00
kind of plugging had all that all others you need
00:15:03
your lens lookup system could also use reporting
00:15:09
i would say the extent is usually where the funding comes from whether it's
00:15:14
a uh oh private uh investor already whether or that that one's true
00:15:20
you start making a product or if it's a of public interest then i i those
00:15:25
things can i think it can not i it's not hard to head into
00:15:29
you need years of looking or something like that and be distributed but it's really about uh whether we will be allowed or not
00:15:36
and okay generally i i'll send him i think he's a year
00:15:41
i challenge all the people want to develop their own application
00:15:45
uh_huh is the same problem we had was on the lens is to what extent there's a community of developers in switzerland
00:15:50
and to what extent we can make it easier to share new content so
00:15:54
that we can do at lowe's faster these results things are um
00:16:01
it really depends on uh i have a i think on and
00:16:04
and who explained on this case because whether the the the
00:16:09
just if the fine it's really depends on where the funding is coming from
00:16:13
because if they require we we cannot uh sure it we cannot
00:16:16
do anything but if it's for us usually we would like to be
00:16:19
able to share and should discuss uh exactly how we did things
00:16:23
uh like for the calibration we spend actually quite a long time to get it quite twelve
00:16:28
so that we can find that transformation in a somewhat easy and reliable reliable way
00:16:34
but we can not a share it yet for instance
00:16:41
but you
00:16:44
local book is still

Share this talk: 


Conference Program

Mot d'introduction
Sébastien Mabillard, Coordinateur projets e-health, Fondation The Ark
8 June 2017 · 6:05 p.m.
265 views
Les Technologies Cognitives ou l'Homme Augmenté
Jérôme De Nomazy, IBM Suisse
8 June 2017 · 6:09 p.m.
Les Technologies Cognitives ou l'Homme Augmenté - Questions du public
Jérôme De Nomazy, IBM Suisse
8 June 2017 · 6:41 p.m.
Mot de bienvenue
Sébastien Mabillard, Coordinateur projets e-health, Fondation The Ark
9 June 2017 · 9:04 a.m.
Présentation de Swiss Digital Health
Frédéric Bagnoud, CimArk
9 June 2017 · 9:07 a.m.
Présentation des instituts de recherche HES-SO Valais-Wallis
Laurent Sciboz, directeur Institut Informatique de Gestion | HES-SO Valais/Wallis
9 June 2017 · 9:11 a.m.
Virtually Better: How Virtual Reality is Easing Pain, Calming Nerves and Improving Health the Drug-Free Way
Brennan Spiegel, Director of Health Services Research, Cedars-Sinai Health System
9 June 2017 · 9:21 a.m.
112 views
Augmented Reality-assisted surgery
Luc Soler, Directeur de la R&D et Professeur PAST, IRCAD, IHU
9 June 2017 · 11:22 a.m.
116 views
Réalité virtuelle dans le traitement des troubles mentaux
Eric Malbos, Psychiatre et spécialiste du traitement en réalité virtuelle, CHU Conception
9 June 2017 · 11:56 a.m.
143 views
La Simulation Numérique pour les Professionnels de Santé
Xavier Abadie, Directeur du Développement International, SimforHealth
9 June 2017 · 12:28 p.m.
AR for different medical use cases
Antoine Widmer, Professeur HES & co-fondateur d'Adventures Lab
9 June 2017 · 2:04 p.m.
Augmented reality for medical applications
Henrique Galvan Debarba, Chercheur senior, Fondation Artanim
9 June 2017 · 2:15 p.m.
Virtual reality solutions for stroke rehabilitation
Andrea Serino, Responsable de la recherche en neurosciences, Mindmaze
9 June 2017 · 2:34 p.m.
Virtual Reality combined with haptic robotics for rehabilitation
Aurélien Fauquex, Directeur Exécutif et co-fondateur, Lambda Health System
9 June 2017 · 2:48 p.m.
Creating a valuable user experience for nursing home inhabitants
Roel Smolders, Directeur Exécutif, Active84Health
9 June 2017 · 3:05 p.m.
TABLE RONDE : Quelle innovation pour la santé du futur en Suisse?
Eric Beer, Henning Müller, Pascal Detemmerman, Victor Fournier, Resp: Head of Customer Engagement e-Health (Swiss Post), Professeur en informatique médicale HES-SO Valais-Wallis, Directeur Exécutif SwissEnov, Chef du service de la santé publique du Canton du Valais
9 June 2017 · 3:20 p.m.