Apply HN: Visage – Fully Automated Makeup Application | Hacker News Reader

Apply HN: Visage – Fully Automated Makeup Application

4 pointsjoyeuse670114 comments

The makeup industry is ripe for change. Currently makeup is being applied the same way it always has - with brush, pen and applicator pad. Now imagine a device that automatically applies makeup. Cut makeup time from 30 minutes to 30 seconds. Users will be able to choose from thousands of colors, shades and products for their ultimate customized look. Makeup artists can create and sell digitized looks for their fans to apply via a digital makeup store. Don't like a look? wipe it off and try another instantly. Having a tough time getting those eye wings right? Try with computer precision down to the millimeter! Tired of waiting while your partner takes 2 hours in the bathroom? They’ll still take two hours, but with this product, they'll have tried dozens of looks instead of struggling with one.

How do we do this? Facial recognition software, reverse kinematics, airbrushes and custom applicators. Our software will analyze a user’s face using proprietary algorithms, enabling a preview of the applied makeup customized specifically to the individual user’s features via our smartphone app. A user may then edit and fine tune their look to perfection and store it on their device for later use - and even share it with their friends!

We firmly believe that this idea is both technologically feasible and inherently marketable. With strong community support and the help of industry experts, we intend to challenge the traditional methods of cosmetics application while placing the power of precision styling in the consumer’s hands, all at a fraction of the time and cost of conventional techniques.

Is the physical manifestation of the product a "robotic arm" (to physically apply makeup) accompanied by software?

In terms of makeup artists selling digitized looks to their fans via a digital makeup store: makeup already happens for free on youtube and is lucrative (per youtube payment of some ad revenue) only for those with large followings. So, people are unlikely to want to pay for a look they can access freely elsewhere.

While the tech may exist, this would require long incubation period for building. I.e. the face has different surfaces (lips, eyelid, brows, etc) that interface with various motion types (glide, etch, press, pat, etc), with each area requiring a combination of motions. Different faces also require different amounts of product+motion. So, to devise the making up of one feature alone (perfecting applicator's motion range, pressure, adjustability to different head/face shapes, etc) would take a while.

It'd be also interesting to see how cost intersects with price acceptable to interested consumers.

In some of the responses below I mention the design to be something akin to an airbrush printer. A half-cylinder shaped rail guided system that allows for full articulation around facial features.

Addressing makeup: People are unlikely to want to pay for something they can get for free elsewhere is not necessarily true. We know this given that the majority of people will buy an OS instead of using a free one like linux, or using alternative app stores instead of gatekeeping software such as the mac app store or Google play. Makeup artists deliver free tutorials, suggestions and tips. To replicate the look requires purchasing the same products and applying the makeup with the same skill level as the makeup artist. If a look to own is $0.99 and can be instantly applied in less than a minute, we believe this stands to be an attractive option.

We believe we can do this better. Say you want a painting in your room, instead of watching a Bob Ross video on how to paint a particular landscape, he gives you a digital copy you can print at home. In a few minutes you can have an excellent high quality print adorning your wall, or you could go through the process of doing it yourself.

Incubation period for a fully featured product would be long, but for argument's sake, developing an MVP that allows for foundation, eyeshadow, lips, brows, blush would be doable via airbrush alone (as it currently is by hand) in a reasonable period of time. Other motion types with different applicators would require further study and consideration. You raise a good point that different faces would require different amounts of product. This may be adjusted by the end-user or heuristics applied when facial data is scanned in.

Current professional airbrush kits are approximately $100-$500 in range. An initial, albeit lofty goal, would be to deliver a subsidized main device sub $1000 USD, with purchasable makeup refills as main revenue.

I see airbrush being applicable to application of dry materials like face powder or eyeshadow. But application for "wet" items like lipgloss, liquid foundation (especially for dewy finish) or liquid eyeliner indicate tooling for equipment that can handle wet and dry materials and cleaning (or designing for residue elimination to be non-issue) incorporated for the different materials being handled.

I can see users agreeing to purchase a look if that means actually a download of instructions by robot - that wasn't clear previously. In that case I'd plan for diminishing returns in revenue here because like the Nespresso-Keurig-generic coffee pod DRM wars, users will eventually find cracked versions of new looks (from elsewhere online, other users, etc) to use.

Current airbrush makeup technologies are meant for wet materials, but you raise an interesting point, that for versatility wet and dry options would be useful and would require either separate tools are a special means of using both and handling cleaning.p

Yes, essentially the looks offered are digital instructions for this makeup applicator. One would ideally model the digital store similar to Apple, effectively acting as gate keeper to software for this technology. The makeup refills could very well be replaced by generic cheaper refills. To prevent substantial drop in revenue would require a moat of sorts, be it brand loyalty, DRM, or strategic partnerships with existing makeup brands to take advantage of existing brand loyalty when it comes to refills.

I can totally understand the software part and I think thats feasible with our current level of technology.

The hardware seems like the biggest challenge. You would need very precise robotic arms and spray-brushes to actually accomplish application of makeup. Although I'm no makeup expert, it seems precision is important (See https://www.youtube.com/watch?v=WcW70-6eQcY).

It's a great video =). Our design is trending towards more of a printer design, with airbrush head(s) attached to curved track able to maneuver and angle around the length and width of a face. The precision to apply foundation will differ from say, eyeliner, but we are confident that the great technological strides in printing technologies would allow us for this accuracy.

Having said that, the other design I think you're suggesting, a few very precise robotic arms is equally valid and possible given the strides in robotics within the manufacturing industry. A consideration to this design would be anchoring the head. For the arms to accurately apply makeup would require a user to remain still, so a head rest may be necessary/helpful.

This surely can't do eyeliner or mascara (too precise). Foundation or lipstick require material that I don't think can be airspread. I mean same as others, if you can pull it off, that's great, but I seriously doubt it.

Maybe for kids...

We believe it certainly is possible to apply eyeliner and mascara. Consider high resolution printing. Your common inkjet printer has sub-millimeter precision. Sure, this is a 3D application but we already have 3D printers down to the micron!

Airbrush makeup has existed for many years and is used typically in professional applications. A cursory search of google yields multiple products by leading brands offering this solution,and it has become increasingly popular with the advent of high resolution broadcasts. Please see: http://www.sephora.com/perfection-mist-airbrush-foundation-P...

http://www.luminessair.com/homepage.aspx

Lipsticks have traditionally come in a solid form but liquid lipsticks exist and are popular and we believe to be useful in an airbrush application:

http://www.airbrushmakeup.com/makeup/liquidlipsticks/

This is a really hard problem, do you and your team have the necessary skills to pull it off?

How will the makeup actually be applied?

Yes, it is a challenging problem, but we figured that's why it may attract y-combinator, a shot to the moon. I am a software engineer and believe it within my skill set to personally handle some of the challenges of the product. From a high level view there is:

3d-scanner (hardware & software component):

At this level developing or incorporating a 3d scanner to obtain precision data of a user's features.

Make-up application data (software): This is the heart of the application. It is essentially printing data in the form of: makeup layer, position (x,y,z), applicator angle, color, opacity and flow.

This requires software for it's creation, previewing and modification and distribution (smartphone app ideally) as well.

Printer software & hardware:

So the actual hardware of an airbrush/applicator with precise x,y,z positioning, angle, and opacity and flow control. The airbrush itself has to consider problems of cleaning much like inkjet technologies had to consider.

The software is of course interpreting the makeup application data resulting in hardware actualizing it. I do not have experience in this field, but I believe it is within reason to form a team of engineers that does, likely from the 3d printing industry.

I hope that helps paint a better picture of what we'd like to accomplish.

You are seriously underestimating the difficulty of producing the hardware you're envisioning.

Have you gotten input from roboticists and industrial designers?

Consider applying lipstick. You'll need a real-time high resolution scan of the person's face (you can't just take a single scan and be done with it, a human is never still), be able to accurately find the lips and their borders, have an articulated robotic arm, grasp the lipstick, have enough of the lipstick out to apply without breaking, apply the right amount of pressure, and finally follow a constantly moving contour on a person's face.

You could probably do blush with an airbrush, but you couldn't do a base or concealer because you'll have to brush or smear that on and you're back to following contours with an articulated arm.

I just don't believe the necessary technology exists for this to do it in a consumer device for less than several hundred thousand dollars.

Perhaps you're right. Though I think the hardware you may be envisioning is different from ours. In the consideration for lipstick application, this device isn't meant to be a robotic copy of a human arm applying various cosmetics a la the jettisons. The goal is to provide makeup application: general cosmetics to specific regions of the face in a given order. From this you can develop cosmetics that can be applied via airbrush and perhaps another applicator (gel, or inkjet) that render this effect. One does not need to stick to conventional solid lipstick nor it's means of application to achieve the desired look of red upon lips. Granted there is mention of kinematics which implies robotic arms and that design path is still a consideration, but recently we have trended towards a means of application via a moveable applicator head attached to rails and guides similar to 3D printing.

As provided in previous links, airbrush is already used for applying foundation, lips, eyeshadow and blush. https://en.wikipedia.org/wiki/Airbrush_makeup

This could be quite cool if technologically feasible, do you have a prototype or how far along is it?

We currently do not have a prototype, the design is in it's conceptual stage.

Reply on news.ycombinator.com