- il y a 6 heures
New Frontiers of Beauty Powered by Tech by L’Oréal Groupe
Catégorie
🤖
TechnologieTranscription
00:02Hello, hello.
00:04Hello.
00:05Hello, hi everyone.
00:06So I'm Laura from Lancome's Beauty Tech team
00:09and today we will present the latest innovation
00:11in our skin tech diagnosis.
00:14So,
00:16we introduced Cell Bioprint.
00:18It's our latest innovation in skin tech diagnosis
00:21and for the first time,
00:22we look at your skin biology
00:24to understand how your skin is aging
00:26and which factors are impacting it.
00:30So,
00:31Cell Bioprint is the first biological skin diagnosis
00:34that for the first time goes beyond skin surface
00:37to look at what's invisible.
00:39So,
00:39how do we do this?
00:42Our device is based on a patented lab-on-a-chip technology.
00:46This means we use a microfluidic technology
00:49to analyze our protein biomarkers.
00:51So,
00:52which are protein biomarkers?
00:54These five biomarkers are each linked
00:56to different biological functions of the skin.
00:58So,
00:59we have biomarkers that concern skin inflammation,
01:02some that look at skin barrier.
01:03And through these biomarkers,
01:04we can look at your unique genetic factors
01:07that impact your aging path.
01:10So,
01:11what is the result that we get with this analysis?
01:14Firstly,
01:14we have the wow factor.
01:16We can reveal the skin biological age
01:18and which stage of biological is your skin.
01:22So,
01:22in this case,
01:23it's reset.
01:23Then we move to step number two
01:26where we can predict your aging signs.
01:28So,
01:29in this way,
01:29we can look at your current skin status.
01:31So,
01:31what's visible and it's your current skin.
01:34And then we look at your future risks.
01:36So,
01:36what we're able to do
01:37is see if you have a particular genetic predisposition
01:41to certain concerns.
01:42That can be wrinkles,
01:43spores,
01:43or skin barrier function.
01:46So,
01:47the step-by-step protocol is very easy.
01:49We will do a simulation right now with Fred.
01:52And as you can see,
01:53it's four easy steps.
01:55So,
01:56great.
01:56Okay.
01:57So,
01:57we'll do a little walkthrough right here with our system.
02:00So,
02:00the first thing you do when you come in
02:02is it is a skin test.
02:04So,
02:04you start with a skin sample.
02:05So,
02:05these are called de-squam tapes.
02:07These are little pieces of tape.
02:08We stick it to each cheek
02:10and we remove an outer layer of the dead skin cells.
02:12And those skin cells have in them
02:14everything you need to read
02:16in terms of the proteins and the biomarkers.
02:18Once I've removed the skin cells,
02:20I put it on,
02:21we call this our protein extraction device.
02:23This is something we had to come up with
02:24because no one has done this for skin before.
02:27So,
02:27we needed a way to get the sample
02:28into our buffer solution.
02:31So,
02:31the next thing is
02:32this container has in it
02:34a small amount of fluid.
02:35It's the buffer solution
02:36that breaks down,
02:37oh,
02:37you can see it right there,
02:38that breaks down all the cells
02:40and all you're left with are proteins.
02:42So,
02:42it's a sample
02:43that has just the suspended proteins.
02:46And the cool thing here is
02:48this is the sample
02:49that we put on our
02:50lab-on-a-chip cartridge.
02:52So,
02:52once this is kind of mixed through,
02:54this item itself
02:56turns into a dropper bottle
02:59and we put one drop
03:00of the sample
03:03beautifully
03:04on the lab-on-a-chip cartridge.
03:06That's all you need.
03:07We call this a lab-on-a-chip cartridge
03:09because,
03:10we use this one,
03:11because the
03:13process normally
03:14to do protein analysis
03:15requires multiple machines,
03:17a full lab setup,
03:18a technician,
03:19a full day.
03:19We do this all the time
03:20in our labs
03:21with our biologists
03:22when we're developing products,
03:23running clinical studies,
03:24et cetera.
03:25What this does
03:26is the cartridge itself
03:27contains antibodies
03:29that bind to each
03:30of the proteins,
03:31fluorescent beads
03:32that bind to the antibodies,
03:33and microfluidics
03:34to fully process the sample
03:36in about 45 seconds
03:37on this trip.
03:38When that's ready,
03:39we put it in the machine
03:40and it shines
03:41a fluorescent laser
03:42over the cartridge,
03:44illuminating the particles,
03:45and that's how we count
03:47the proteins.
03:48So from this,
03:49we get a result
03:49that looks like this.
03:50Should I walk through
03:51the result?
03:52Cool.
03:52So on the bottom
03:53of this screen,
03:54you can see
03:54these are our five proteins,
03:56so the actual levels
03:57that were found
03:57in your skin.
03:58And we know a lot
03:59about what these mean
04:00because of the research
04:01we've done,
04:02but how we present that back
04:03to the user
04:04is as a biological age.
04:05So I can tell you
04:07that your skin
04:07has the biomarkers
04:09of someone five years younger
04:10or five years older.
04:11So it's really
04:12that overall measurement
04:13of the longevity
04:14of your skin,
04:15the health of your skin,
04:16and what's happening underneath.
04:18But like Laura said,
04:19we can go deeper
04:19and our research
04:21has connected
04:21each of these biomarkers
04:22to specific skin signs.
04:24So for example,
04:26there is a biomarker
04:27associated with
04:28barrier functionality.
04:29And what I can do
04:30is I can show
04:31that you have
04:33the biomarker
04:33associated with
04:34barrier dysfunction.
04:36And what that means
04:37is you are more likely
04:38or will eventually
04:39develop an issue here.
04:40And although we have
04:41this other reading here,
04:42by using our imaging device,
04:43I can see what's happening
04:45on the face right now.
04:46Right now,
04:47you have no issue
04:48with barrier function.
04:48So this is that exact case
04:51where we can use
04:51proactive skin care.
04:53We can come in
04:53with a solution
04:54that's gentle,
04:55that's in advance
04:56of the issue ever forming
04:57to prevent it
04:58from ever happening.
04:59So in this case,
05:00some ceramides,
05:01something like this.
05:02Or we can look at cases
05:03where you do have
05:05the biomarker
05:06and you're already
05:07struggling with it.
05:08I see it's an issue already
05:09and this is just something
05:10that you probably will need
05:11a little extra help with,
05:12probably a longevity
05:13oriental solution
05:14specifically for that issue.
05:17Cool.
05:18Well, thank you so much
05:19and if you have
05:20any additional questions,
05:21we wait for you
05:22at the L'Oreal booth.
05:26Okay, thank you very much.
05:27Lancome,
05:28Cell Bioprint.
05:30Thank you very much.
05:31I'll let you
05:34disinstall your stuff.
05:36So I hope you enjoyed
05:38that very exciting demo
05:40from Lancome
05:40Cell Bioprint.
05:43And of course,
05:44if you want to know
05:44more about it,
05:45please,
05:45you can find L'Oreal
05:47at their booth
05:48in the middle of Hall 1.
05:50Very big,
05:51fabulous booth
05:53that it is.
05:54Now,
05:57let's shift our focus
05:58and talk about
06:00the future
06:00of beauty marketing
06:02and creativity
06:03with Gen.AI.
06:06So I would like
06:07to welcome on stage
06:09a L'Oreal
06:10and Gen.AI
06:12artistic duo,
06:14Thomas Alves Machado
06:16and Laszlo Gao
06:19who are here
06:20to introduce
06:21Crea AI Tech.
06:46Hello, hello.
06:47Hello, no.
06:47Ah, hi everyone.
06:49So I'm Thomas Alves Machado
06:51and I am the head
06:53of the Gen.AI
06:55content program
06:56at L'Oreal
06:57that we call Crea Tech.
07:00So Gen.AI
07:01goes at the speed of light
07:02and as you know,
07:05for marketers
07:05it's a key challenge
07:07to keep up
07:07with these innovations.
07:09Agility is the name
07:11of the game,
07:12sorry,
07:13and as we say
07:13at L'Oreal,
07:15we like to size
07:17what's starting
07:17because we are
07:18a beauty tech company
07:19with more than
07:20100 years
07:21of beauty history
07:22and it all started
07:24with innovations.
07:26So more or less
07:27two years ago
07:27when Gen.AI
07:28heated the world
07:29we saw the opportunity
07:30to mix creativity
07:31and technology
07:32and that's how
07:34we create our program
07:36which is our safe ecosystem
07:38for brand experimentations.
07:40We're exploring first
07:41all those new technologies,
07:43testing them,
07:44producing text,
07:45context, image
07:45and then
07:46we develop our platform
07:48where we integrate
07:48all those new technology
07:50for adoption at scale.
07:52But more than that
07:53we do it
07:56sorry,
07:57more than that
07:57we learn
07:57and we do it
07:58standing by our commitments
08:00because we do not use
08:01Gen.AI
08:02to generate
08:03lifelike face,
08:05body,
08:05hair
08:06or skin
08:07to support
08:07and enhance product benefit
08:08in our external communications.
08:10So what does it look like?
08:12So you can come
08:13see us in the booth
08:14but we do brand custom model
08:15to ensure brand consistency.
08:17We have also 3D digital twins
08:19to control the rendering
08:20of the outputs
08:21and image to video
08:22to stretch creativity
08:23to the limit
08:24and this is without a doubt
08:25the next opportunity
08:27to scale for us.
08:29Through our strong relationship
08:30with Google
08:31we use this technology
08:33like VIO
08:34to create
08:35brand video content
08:36and we see
08:37those technologies
08:37not only
08:38to enable us
08:39to create
08:40high quality content
08:41quickly and at scale
08:42but it opened up
08:43a new door
08:45of creativity.
08:46Good tech
08:47in the hands
08:47of talented human creative
08:49will lead to more
08:50impactful content
08:52and that's what
08:52we're going to see
08:53today with Laszlo.
08:54What happens
08:55when you blend
08:55an iconic brand
08:56like Lancôme
08:57a talented
08:58the talent
08:59of the greatest
09:00AI artists
09:01and Google
09:03without technology
09:04let's see.
09:06Hello everyone
09:06my name is Laszlo Gala
09:08and I've been working
09:08in post-production
09:09for the past 20 years
09:10but of course
09:11in the past two years
09:12I'm experimenting
09:13with generative AI
09:14mainly bringing
09:14cinematic pieces
09:15to brands
09:16as commercials.
09:17So here today
09:18what we will see
09:19is that I will
09:20try to show you
09:21how creators like me
09:22can leverage
09:23generative AI
09:24to enhance
09:25their process
09:25and also
09:26that enabled us
09:27to create
09:28the end result
09:29in less than a week.
09:30So when we're talking
09:31about any kind of
09:32film or video production
09:33we can talk about
09:34three phases
09:35pre-production
09:36production
09:37and post-production.
09:39During the pre-production
09:40phase
09:41you receive the brief
09:42from the client
09:43and you have to check
09:44how to solve it
09:46and of course
09:46at that moment
09:48you can use LLMs
09:49train them
09:50fine tune them
09:51so the visual identity
09:53and the brand identity
09:55will stay consistent
09:56throughout all your prompts
09:58all your generations.
10:00Solving the client brief
10:01itself
10:01is a very very creative process
10:03so that's still up
10:05for the human
10:05but you are in the driver's seat
10:06and the AI will enhance
10:08and make your process
10:10much easier.
10:11So for the pre-production
10:12we can use LLMs
10:14to get to know
10:15for example
10:15the brand better.
10:16Here you can see
10:17I'm just asking
10:18Gemini
10:18to explain
10:20why is there a rose
10:20in Lancome's logo
10:21and it gave me a history
10:23of the brand's past.
10:27The brief from Lancome
10:28was to create
10:30a visually delicious campaign
10:31for the Juicy Tubes
10:33coffee range
10:33but with the challenge
10:35that we cannot include
10:36AI generated humans
10:38so that's a little bit hard
10:40but how we solved it
10:41is that we created
10:43very very interesting
10:44camera movements
10:45with the textures
10:46of the products themselves
10:47so for this mostly
10:48I used text video prompting
10:51and you can see
10:52that you can input
10:53one image of the product
10:55and Gemini will help you
10:57to translate it
10:58into a text
10:59that later on
11:00VO will understand.
11:02or another simple use
11:03for LLMs
11:04is that we wanted
11:05to use nostalgia items
11:06from the 90s
11:07so Gemini was able
11:09to help me
11:10to bring a list
11:11a numbered list
11:12and later on
11:13I can just reference
11:14the number
11:14and it will include it
11:15in the prompts.
11:16So LLMs can also
11:18help us to create
11:19the prompts themselves
11:20I will show you
11:20a quick example
11:21a short example
11:22but you have to know
11:23that what you will see
11:24for one video
11:26so one little clip
11:27sometimes I used prompts
11:29that were as long
11:30or more long
11:31than my current speech
11:32so again
11:33not for the whole video
11:34but just for one clip
11:36sometimes I used prompts
11:37that were longer
11:38than my current speech
11:39so the last step
11:40of the pre-production
11:41is to create
11:42the workflow itself
11:43because each project
11:44needs a different
11:45kind of workflow
11:45we were talking
11:46about text to video
11:47we were talking
11:48about image to video
11:49we can also use ingredients
11:51to enhance object
11:53character
11:54or scene consistency
11:55or we can also use
11:57the extend feature
11:58which can generate
12:01with a similar camera movement
12:03you can continue
12:03your previous generations
12:05in a more longer shot
12:06for these extra features
12:08we can select the models
12:09that enables us
12:11to use these features
12:12with different kind
12:13of speed
12:15and inference speed
12:16and of course cost
12:17and you can see
12:18through this creative
12:19interface
12:20we can select
12:21the model that we want
12:22for the product shots
12:23we use an image
12:25to video workflow
12:25why?
12:26because Create-Act
12:27already had
12:28the 3D digital twin technology
12:30which means
12:31with the general model
12:31for example
12:32with mid-journey
12:33it's extremely hard
12:34to get consistent results
12:35and text, logo
12:37and product accuracy
12:38but with the 3D digital twins
12:40everything was solved
12:42so now you have
12:44some kind of idea
12:45how the individual shots
12:46are built
12:46how can we make
12:47a more consistent output
12:49with the help
12:49of color grading
12:50what is color grading?
12:51we just try to match
12:53the generated output colors
12:55more with the brand's colors
12:57here I'm using
12:57the DaVinci Resolve
12:58with a little tool
12:59on the bottom right
13:01called the vectorscope
13:02to match the brand's colors
13:04of course
13:05everything that we generate
13:07sits on our timeline
13:08so horizontally
13:09we are building our story
13:10but vertically
13:11we can build our versions
13:13it can be previous generations
13:15or it can be new generations
13:16so the iterations
13:17with the clients
13:18will be much
13:19much faster
13:20so when everything
13:22is ready
13:22and the client
13:23is happy
13:23then we can export
13:24and let's watch the results
13:53and let's watch the results
13:54down
13:54and let's watch the progress
13:54in the bottom
13:55and he Grow
13:55just to catch up
13:55and then he'll be
13:56Sous-titrage Société Radio-Canada
14:26Thank you. Wow. So, thank you guys. You just gave us a stunning glimpse of the future of beauty marketing
14:31and creativity with Gen.AI. Very, very nice. Again, if you want to go over to the booth, L'Oreal
14:38booth, a little bit later, you can unleash your own creativity and test and prompt your own visuals.
14:47All right. Now, let's shift gear and dive into the future of e-commerce with our last initiative with Noli
15:00.com.
15:00The first AI-powered, 100% personalized online beauty shop, backed by L'Oreal Group.
15:09I welcome on stage, please, please welcome, Amos Suskind and Mayel Gask, CEO and Deputy CEO of Noli.
15:31Hello, everyone. Have you ever felt lost when trying to find the right beauty products? I'm sure you have.
15:40Actually, 90% of the beauty products are not a perfect match for you.
15:49My name is Amos Suskind. I'm the CEO of Noli and I'm joined by my co-founder.
15:54I'm Mayel Gask, the Deputy CEO of Noli.
15:56And together with our founding partner, L'Oreal Group, we are very excited to solve the biggest pain point in
16:02the beauty industry, which is helping people find the right beauty products.
16:08If we take a step back, I'm in the beauty industry for 20 years and more, actually.
16:13And beauty is supposed to be driving joy, the joy of discovering, the joy of self-expression.
16:21It's a beautiful market that's growing fast.
16:24By the way, the e-commerce beauty market is supposed to grow double-digit in the coming year.
16:28So it's a very positive, exciting playground.
16:31But there is a big paradox.
16:34Shoppers actually feel completely lost and overwhelmed.
16:38They feel frustrated because, in most of the cases, they buy the wrong product.
16:43And down the line, they're actually deceived.
16:46In the UK only, every year, one billion pounds worth of beauty products are wasted because they are not addressed
16:53to the right person.
16:56100% of beauty shoppers and beauty lovers actually say that they struggle to find the right skincare product.
17:03This is a big problem.
17:05That's why we are very excited to launch Noli, which is our cutting-edge tech to solve the biggest pain
17:11point in beauty.
17:32So, with Noli, Noli.com, we are building the first AI-powered, 100% personalized online beauty shop.
17:41So, with Noli, Noli.com, we are building the first AI-powered, 100% personalized online beauty shop.
17:42The brand stands for No One Like I, which is our tribute to personalization.
17:48Maëlle, could you explain us more?
17:50Thank you, Amos.
17:52So, what is a 100% personalized beauty shop?
17:55Well, it starts by you, by decoding yourself.
17:57So, we do a diagnosis.
17:59It takes two minutes.
18:00It's a combination of questions and face scan using computer vision.
18:04Let's do mine together.
18:06So, I go on Noli.com.
18:08By the way, all of you can do it from all over the world.
18:11And you answer a few questions.
18:13So, I'm answering, giving my date of birth, gender, and then I'm doing the face scan.
18:19This is a picture of me.
18:21The algorithm analyzes a lot of different data points and is trained on 1 million skin data.
18:30It gives an accuracy, a diagnostic that is as accurate as 98% versus a dermatologist.
18:38And here, it recommends focus areas for me.
18:40So, it went a bit fast.
18:41My focus areas are fine lines and wrinkles, blemish prone skin, and also dark circles because tough startup life.
18:48Then, I'm giving some more information about my budget, my scent preferences because Noli advises your product that you will
18:56love also.
18:58And the result is a completely bespoke and personalized routine for me.
19:05And not only is it recommending products, it's recommending alternatives and explaining why I get recommended this product.
19:12There is no retail media on Noli.
19:15There is no us pushing brands that we favor.
19:17It's 100% unbiased, backed by science, and we're live in the UK, and we send the routine at home.
19:25So, this is my routine that I receive in 24 hours at home, thanks to Noli.
19:31If you don't want to shop the full routine, you can also browse the website, and each section is completely
19:37personalized to you.
19:38So, if you go and shop all or in CeraVe or Lauche-Posay, they're your favorite brands, the products that
19:43you will see on top are the ones that are good for you.
19:46So, this is my routine.
19:48I won't go into details, but please come to our booth to discover yours and understand why this is a
19:52great routine for fine lines and wrinkles, acne, and dark circles.
19:57Four things to remember about Noli.
20:00First, we're user-centric, unbiased.
20:02This is the first time an e-com shop is really focused on the user and helping them solve their
20:09problem before everything.
20:11Second, we're fueled by the science of L'Oreal.
20:13Our great partner is giving us access to their unlimited scientific knowledge and user insights that we use in our
20:22algorithm.
20:23Third, we have 21 brands live and we're expanding rapidly.
20:27So, you can shop brands like La Roche-Posay, Kiehl's, Lancôme, and much more.
20:32And finally, we're leveraging cutting edge tech and AI because we were born just a year ago and we have
20:37the chance to be AI native.
20:39We've built our platform around that.
20:43And we have started our adventure in the UK, in London, and the results so far, the feedback that we're
20:49receiving is mind-blowing, super exciting for us.
20:51We are a startup with a purpose.
20:53We want to solve the biggest pain point in the beauty industry.
20:56We want to change people's lives when they buy beauty products.
20:58And the feedback is really amazing.
21:00We have a lot of testimony.
21:01You can see some of them on the screen that really confirm that finally they find product that they love,
21:08that finally they can buy product with confidence.
21:11And for us, it's a fantastic tribute to the success of our team.
21:14And actually, it motivates our team a lot, showing that with AI, we can actually deliver genuine value for the
21:20end user.
21:22So if you want to know more, please join us on our booth.
21:24We are hosted by our L'Oreal backers.
21:27So we are backed by L'Oreal Group.
21:28So we have a fantastic booth over there.
21:30Please pass by the booth, do your beauty DNA test, receive your routine recommendation.
21:35It's free for everyone.
21:37And if you go and you are lucky, one out of five actually received one of the serum products.
21:41So go there and good luck.
21:43Thank you very much.
21:44Thank you.
Commentaires