Facebook PixelWhat If You Had Your Own Personal Stylist?
Create newCreate new
Sessions onlySessions only
Ideas onlyIdeas only

What If You Had Your Own Personal Stylist?

Daami A.
Daami A. May 06, 2021
Here's the quick idea.
Imagine you have your designated personal stylist. Your stylist is an app - an app capable of casting to a television screen. Whatever works for you, the latter part is not necessary, you might be okay with seeing yourself just on a smart phone with no larger screen acting as a faux mirror.

What does the app do?
The app is there to "find" you your style. Your brand. Style and branding is intentional, and creating an aware unique style is one of the surest way to land a good first impression. Impression matters. Impression can land you your job, Can land you a spouse. Can create opportnities you did not know existed. Asides that, a good style is cool!

So how does the app work?
The app sweeps the latest trend of clothing styles and combinations on the internet, especially on image-oriented sites such as Instagram and Pinterest. There are different names or categories for a group of styles. If the algorithm has a solid unsupervised learning framework, or a semi-supervised one (whatever works in this case), then we could have a list of categories before we begin browsing through combinations.

This way, first, a user can choose between a plethora of common categories such as Minimalist, Retro, Old Fashioned, Chic., Bohemian, Sexy, Artsy, Casual etc, etc.
Once a category is chosen, the user can go through the combinations suggested to pick a style to their taste. The AI stylist tries different available combinations of styles on the user's full portrait so the user can see what style suits them best.

To give the user more control, they might be able to play with a lot of features and accessories to see what works for them or matches their skin tone, profile and body type.
More control options might.be available, like options to change colors for clothes with plain colors, or options to experiment with hairstyles.

AI has come a long way but this would require some really good advancement in technology. I doubt this would be achievable the way I'd love it for be: perfect for its intended audience. There are few existing underwhelming versions of this, and even then, most of them are more concerned about the existing clothes in your closet. Doesn't exactly work for my required type of audience. Especially when you could just as easily try on those clothes instead of taking pictures of them to see if they fit!

Who is the App for?
The app would be perfect for:
  • People with zero to no fashion sense
  • People looking to explore
  • And lastly, for people like me, who quite have the money to get good clothes, but don't know the right clothes to buy.
Instead of getting a costly stylist, why not get a personal one that gives you a chance to explore tons of options without pressure?
Creative contributions

A community to help you select the best version of your style by voting

Darko Savic
Darko Savic May 06, 2021
The app would have a "help others decide" section where a bunch of side-by-side options would show up. People would click on whichever they think looks better. After that a new set shows up from a completely different person, and so on.

From your perspective, you would select 2 of the options you can't decide between and put them up for voting. The system would run it by 100 people really quickly and you would have your answer within minutes.

Different demographics prefer different styles. You would be able to set the preferences based on the age and location of the people who should be asked to vote.
Juran4 months ago
It's a really cool idea and a really cool contribution! I agree with everything you posted, but I just want to raise one concern. Would it be your style if it's chosen by your followers? Does "your style" needs to be a product of your find&match process, or it's enough just to be worn by you to be yours? :D I would be worried that I would lose myself somewhere along the way...
Daami A.
Daami A.4 months ago
This is a very good suggestion. I love that you can choose preferences and decide the type of demographic you want to consider.
And putting a style up for voting is even more brilliant because it relies on the user's choice. Some do not want the opinion of others, some are guarded against it.
However, if it worked with the ubiquitous "follower-following" pattern of most social media apps, then that would be better, because then you will be sharing your portrait between people you trust and whose opinions you value.
Throwing out your portrait to a bunch of strangers who can comment and be unkind might be counterproductive to the app's original purposes, after all, the aim is to find your own style.

Again, all these should rely on choices. This is the main key here. A user could choose to share a portrait to 1. Their followers/friends 2. A particular demographic or community 3. The public

Quite brilliant!

I believe what would make even more sense is that the app would offer one of the best platforms to create a solid portfolio as a fashion stylist. Stylists have all the tools at their disposal to design complete outfits, shoes, hats, everything, for their followers.

if you can attract influencers along with your target audience, then you have the perfect social media app already.
Martina Pesce
Martina Pesce4 months ago
Would also be cool to be able to weigh the importance of the vote.
Maybe you can give more importance to a user you like the style of or that have already gave suggestions you liked.

3D scanning for clothing fit

Manel Lladó Santaeularia
Manel Lladó Santaeularia May 06, 2021
Thank you for sharing this great idea, I think it has a lot of potential. I wanted to build a bit on it by proposing that the app incorporates a 3D scanning system that allows you to scan your own body. This kind of 3D scanning is already being used for fitting clothes better for customers. While the 3D scan method used in the link I shared is definitely not translatable to an app that you can use at home, there are really good softwares nowadays that allow us to 3D scan objects and people and quickly have a precise 3D representation of them, which can then be used by different programs. In particular, the 3D body app takes two pictures, one from the front and one from the back, and uses them to create a 3D avatar of your body from which it can take a hundred different measurements, including Body volume, Cervical height, Crotch height, Mid neck girth, Shoulder width, Shoulder length, Shoulder breadth, Bust/chest girth, Underbust girth, Waist girth, Hip girth, Arm length, Upper arm girth, Wrist girth, Max thigh girth and plenty of others.

Coupling such a system with a clothing app would be amazing. Imagine having a visual representation of yourself, with your exact measurements, in which you can try any clothes you want. If this was really well done, the app would have a functionality similar to that of flight search engines, which agglutinate results from different company webpages. In this case, the app could include clothing from different brands (the more the better) and allow you to select based on price range, style and fit. Once you selected a piece, you could put it on your 3D avatar and see how it fits you. This would be the end of one of the main problems with ordering clothes online, which is the fear they won't fit properly. Even if companies tend to give measurements of their clothes in centimeters or inches, a lot of times they are not accurate, or you fear the clothes will not look good with your particular body shape. This would solve the issue by showing you exactly how the clothes should fit on you. Of course this would need a lot of tweaking and collaboration from the brands, but I believe it would be extremely beneficial for them long-term. The other reason while I think this would be amazing is because you could try on and combine clothes from different brands to create your own unique style. Maybe you like a particular set of jeans from brand A but you like t-shirts from brand B. You could literally try the combination on and decide if it looks good before you buy it.

Can we come up with ways to facilitate/integrate the involvement of different brands and the way to generate an economic benefit for the app? Would a subscription system work or would it be better to ask the brands for a percentage of sales?

Martina Pesce
Martina Pesce4 months ago
The best implementation to a perfect brilliant idea.

I think it's the most suitable historical period for such an invention to easily involve brands. Maybe you could actually try to target mainly small brands. It would be great since it gives a chance to a market which is slowly dying because of covid considering they are struggling to stay open. This way there is no need for shops to be open and they would easily manage to extend their client base throughout the Internet without the difficulty of "Oh! Are these clothes going to fit me? ".
I believe involving brands is gonna be easy! They are really waiting for an idea like that!

At least for the beginning, it may be better to ask brands a percentage of the profit. Once the app will gain some notoriety, a monthly subscription could also be an idea. I think it also depends on the targeted clients. Which depends on the kind of available clothes. Which depends on the brands on board.
I think it's good to start with a percentage on brand and adjust the strategy along the way.

So excited to see this project starting!
Ideas for the name of the app? :)
Manel Lladó Santaeularia
Manel Lladó Santaeularia4 months ago
Hi Martina Pesce thanks for your comment. I agree that we are in the perfect time to launch an app of this sort. It would solve the issue that stops a lot of people from buying clothes online: not knowing if they will fit. However, I'm worried about the reception this kind of app would have from major clothes brands. If you think about it, most brands have their own apps where clothes are shown on the best possible models to show how good the clothes "can look". This is radically different to showing how they look on you. Additionally, there could be two main problems to get brands involved:

- You need their collaboration in measuring every piece of clothing in several ways, maybe even scan it. And not only one size of it, but all of the sizes of each clothing piece. That is a very big time and money investment that goes away after you remove that collection. Will it be worth it for them?

-The idea of the app is to integrate several brands in order to be able to compare or even combine clothes from them. Would the brands be on board with this or would they be scared to comparatively lose to the competence? I'm concerned that without including the most popular and accessible brands, the app would not attract the same amount of users and such would not be as profitable. Can we think of ways to make the whole thing more attractive to the brands?

Manel Lladó Santaeularia
Manel Lladó Santaeularia4 months ago
Martina Pesce As for the names for the app, that could be a difficult one, let's try to come up with a list of ideas:

-AboutTheFit (this one may have a trademark since it's from the movie "The Intern"
-The Dressing Room
-The Fitting Room

Let's see what else we can come up with!

How to create an AI that can give you proper style advice and listen to you

Manel Lladó Santaeularia
Manel Lladó Santaeularia May 06, 2021
In this contribution I would like to touch on the practical side of making an AI that could act like a real stylist would do. For this, I'm going to base my ideas on the AI and feedback system used by the personal training app Freeletics which I have used extensively for the last 6 months and which I believe has implemented a very smart and successful approach. In this case, obviously, the aim of the AI is completely different, but its functionality is translatable. Because the Freeletics AI is trying to act as your personal trainer. And how does it do that? This is a step-by-step approach that I will describe in detail:

  • Asking you a lot of questions about yourself: Like any personal trainer would do before they start working with you, the Freeletics personal coach AI starts by asking you a ton of questions about yourself. Some are quite basic, like your age, sex, height and weight, but others are deeper, like "What are your objectives? Do you want to lose weight, gain muscle or just mantain a healthy body as you age?"
  • Letting you set up the conditions you want to work out in: What sense would it make for your personal trainer to ask you to do exercises that involve materials you don't have available? What if you only had half an hour every day but your trainer insisted you have to train for an hour? That would be frustrating and would lead to you not doing all the exercises that are asked of you. To avoid this, the Freeletics app takes into account several conditions that you can imput from the beggining, like "How much space do you have to train?", "What materials do you have?", "How long would you like to train, and how many days a week?" and, importantly, lets you change them later if you want to, and adapts to that.
  • Creating a workout program that takes into account all of those parameters: This is the most difficult part. How do you integrate all that information in a way that can be useful to design the best possible workout program for that particular individual and their goals? Of course a basis of this is having a set of pre-defined plans that can then be tweaked according to the set parameters. However the amount of work involved into making that work is definitely not negligible.
  • Asking for user feedback and adapting to it: This is for me the crucial aspect that makes this AI work beautifully. At the end of every single workout, the app will ask you about it. It will ask you whether you could perform the exercises with correct technique and whether you felt like the workout as a whole was too easy, too hard or had the appropriate intensity. According to your feedback, the app may introduce some easier versions of an exercise you couldn't perform, change workouts to others with higher or lower intensity or even remove exercises if you don't like them.
I know you'll be thinking "Okay Manel, now you've pitched to me why Freeletics is so cool, but what the hell does that have to do with this idea? How do we apply that to a style app?"
Well, I think we can learn a lot from this kind of AI and try to apply it to a style app. For example, starting with introducing your information such as your measurements (or your 3D scan as I mentioned in my previous contribution), your age and the colors and style of clothes you like would be a starting point for the app to begin presenting options to you. You could also set up conditions like the price range you want to look for, brands to include/exclude and the kind of clothing you are looking for (sports clothes, dress clothes, casual clothes) and whether you are open to experimenting with different/new styles or you want to stick to a particular style. After this, the AI could take those parameters and propose a series of clothes to you, which could be categorized using a series of filters or even use a Tinder-like swap system, in which clothes you don't like get directly discarded and the ones you like get stored. You could then, after some selection, compare and combine the clothes you liked. Lastly, the app could take feedback from your selections and discards, as well as from your opinion about the clothes combinations it proposes, and adapt the algorithm to your specific likes and needs. That could be tweaked into the app proposing a style change which could actually better fit your tastes, since sometimes we stick to our same style and are scared to try new things we actually like.

I believe this would be a hard but rewarding project to embark on and that people with the right tools and knowledge could make an amazing app out of this. Can we find ways to establish or improve the ideas I've presented for their applicability into this style app?

Save outfits you like

Spook Louw
Spook Louw May 06, 2021
If Darko's idea of seeing other users' outfits is implemented, you could perhaps have a board, similar to Pinterest where you can save outfits you like.
When uploading your outfit, you can be given the option to include the brand names or the shops where you purchased the individual items.
If you like a certain person's style, you would then be able to follow them in order to see their other outfits. In this way, people could build a following, which immediately creates the opportunity to become an influencer which could lead to sponsorships or endorsements.

The app can be an add-on to Amazon

Shubhankar Kulkarni
Shubhankar Kulkarni May 10, 2021
I am combining ideas from the existing suggestions. Instead of a standalone app, it can be a part of the Amazon shopping experience. This is because once you try an outfit and people vote on it, what then? You will want to buy it. For that, you will require an e-commerce portal and a delivery system. All these other requisites already exist and have proven to satisfy customers globally. So why not create the part that is unique, that is the personal stylist, the 3D scanner that shows you how the clothes fit you, and the voting part, which is the social media part, and create the add-on. Then sell it for an exorbitant price to Amazon. Amazon will buy it in a jiffy and you will be set for life. :)
Daami A.
Daami A.4 months ago
Brilliant! I thought of this too! But not in the capability of selling per se, but a solid integration with the most popular e-commerce sites.
So it is not strictly for Amazon, but an app easily integrable with most e-commerce sites. So this way, perhaps the database you scan for clothing styles is not actually Instagram or Pinterest, but instead, you'll be scanning a plethora of e-commerce sites for your perfect fit. When you find the perfect fit, the price and link to all the items will be available.

There will also be options to check if a particular clothing you'd like to purchase independently fits your body or your style. On this part, however, I think one would need a great deal of collaboration with Amazon or concerned e-commerce sites as well. Because in this case, most of the clothes uploaded would might have to been 3d-scanned by the sellers to achieve optimal results.

Now, how this is possible, I'm not sure.
But industry-changing idea if it's executed.
Shubhankar Kulkarni
Shubhankar Kulkarni4 months ago
Daami A. Defitnitely industry-changing! A lot more in the current crisis.
I went with Amazon since it holds the greatest share of the market, but you are right, it doesn't have to depend on Amazon. However, collaborating with all the e-commerce sites out there may be a lot of work. If we start with Amazon, others might have to follow to stay in the competition.
Manel Lladó Santaeularia
Manel Lladó Santaeularia4 months ago
Shubhankar Kulkarni For me the biggest problem with Amazon is that there are way too many vendors and not all of them are "trustworthy". It would be difficult to make sure that the sizes of the clothes are well reported. Also, as Daami said, it would possibly be necessary to scan the clothes, which is something that small vendors may have trouble with, while larger, more established brands would have an easier time getting into it.

However, I agree that this app should be combined with the possibility to easily buy these clothes, either through Amazon or through direct collaboration with the brands. However, taking advantage of already established and efficient delivery infrastructures like that of Amazon would be game-changing.

Users measure themselves instead of scanning every piece of clothing

Shubhankar Kulkarni
Shubhankar Kulkarni May 12, 2021
@Manel Lladó Santaeularia I see your point. Every piece of clothing will need to be scanned by the store and that is time- and money-consuming.

I propose the following solution: Only the users need to measures themselves perfectly and upload all the values to the app. The app will then suggest 2 (or 3) parameters that the clothes usually display. These parameters are chest circumference in inches, length of the outfit, and/or waist circumference. There can be multiple parameters based on the kind of outfit you are looking at. The app will highlight these three parameters of the user when they decide to try on outfits through the app. The users will decide whether they like the pattern and color of the outfits they wear and post pictures for other users to vote. The images of the outfits from the stores will simply stretch or shrink themselves to properly fit the user. This will only be for visualization purposes and will be done by the software of the app.

The second part is when the user wants to buy the outfits they want. At this point, the app will highlight those 3 parameters and the users will choose what size from the online store will fit them. This is what they do currently. That part would not change. This way it will be less taxing for the brands and they will jump on board relatively easily. Also, the measurement of the users will help them visualize themselves the way they are instead of looking at an outfit worn by models. They may be able to make a better choice because of it.
Manel Lladó Santaeularia
Manel Lladó Santaeularia4 months ago
While I agree that the measurement of the individual is crucial, I have a problem with the fact of using only 2-3 parameters to evaluate the fit. I have ordered clothes following those parameters that were supposed to fit me and, when they arrived, didn't look good on me. That's why I think that using a larger number of parameters would give you a more trustworthy assessment of fit. In this case, scanning the clothes or imput of all the different characteristics from the initial design would be the best two options.

Also, stretching/reducing shapes while keeping them proportional and accurate is not that easy. I think a better system would be to use cloth simulations. There are really cool engines right now that can do amazing cloth simulations pretty easily. With the data of the size and proportions of the body and clothes, and a couple algorithms, it shouldn't be difficult to simulate the fit like that.

Not trying to go against your ideas but I feel that what you propose is way too similar to the existing standard and not revolutionary enough to make this app worth it as an investment. I think it should be something that aims to completely revolutionize online clothes shopping. If you succeed, the brands will adapt.
Shubhankar Kulkarni
Shubhankar Kulkarni4 months ago
Manel Lladó Santaeularia However, as you pointed out, scanning each outfit will be a huge task for the brands and we don't know whether they would be willing to spend so much initially. My idea above has a way around that. I am not suggesting stretching any kind of cloth but the outfits that the brands already have in store. Instead of scanning the outfits, a simple photograph of the outfit will suffice and the algorithm will stretch it to fit the user. Also, the pictures/ photos of all the outfits are already available on online shopping platforms and the brands will not have to spend any more than they do now.

Of course, the original idea of scanning the clothes is much better but has the problem of spending time and money on the scanning part. The current one solves that problem.

The usage of only 2-3 parameters is not for showing how the outfits fit the user. They are for helping the user buy the outfit. And that is the current norm. You don't need to know your bicep circumference and your arm length every time you have to buy an outfit. I agree that we make mistakes and that we may have to return the outfit and exchange it for that of a more appropriate size. But that is not the problem of the new app we are proposing here. That problem already exists and the online shopping platforms have solved it by providing free exchange. So, currently, we use only the above-mentioned 2-3 parameters while buying clothes.

This way we simply install the non-existing technology, which is dressing up virtually and letting people vote on the outfits. All the other tasks are already there. If we just install the new thing and use the existing technology, we can bring the product to the market pretty quickly. After the product is out and people start using it, we can think of making modifications/ algorithms that are going to take longer to build and implement.
Manel Lladó Santaeularia
Manel Lladó Santaeularia4 months ago
Shubhankar Kulkarni I think this boils down to what concept is the most important to you. While you defend that the "social" aspect of the outfit voting is the main one, I would have it as an add-on to the main app, which would be about the clothes fitting and online ordering. That's easily going to give you more revenue since the brands will get involved more easily.

I don't believe that scanning the pieces would be such a problem. Think about the time and money it takes to photograph all the pieces with a model. You have to pay people responsible for makeup, ligth, photograpy, modelling, direction. Invest the same resources into having some sort of "dummy" and camera rig that, when you put some clothes on the dummy, allow you to quickly measure all the proportions. It could even work by integrating touch (from the dummy) and 3D imaging (from the cameras) data to give a more accurate measurement. Once this was set up, you would just have to send one of this systems to each brand so that they try on every piece. If this was well done, the time involved would not be much, and the economic issue would be only to set up/buy the system but would be amortized after some uses. Couple this with a good cloth simulation software, as I said, and you have the best possible modeling of how the clothes will fit you.

While I don't need to know all my sizes every time I buy an outfit, I would only need to scan my body once to have all that information available to the app. And compare it to the pieces I want to buy. I only see advantages in having more information. Also you mention that shopping platforms solve the size problem by providing free exchange. First of all this is not true for all platforms, and secondly this is exactly the point that my idea would try to avoid. Nobody likes the feeling when you've been waiting for some clothes for a while, only to find they don't fit you and you have to send them back and wait again for, hopefully, a better fit. What if that piece of clothing doesn't fit you at all, independently of the size? We all have different bodies. The app I'm proposing would show you how the clothes fit your body so that you would avoid these problems.

Lastly, I wanted to get back to what you said about stretching a photograph. As I already explained, this is not a reliable system if you don't have precise measurements of the clothes. It could be that the app stretches the photo so that the clothes fit me amazingly, and when I buy them that's not the case. What is the use of that then? A cloth simulation, which is easy and cheap to do, with the proper measurements, would work a million times better.

Can we come up with a way to measure the clothes fast and cheap? Some iteration to the dummy+3D camera system I have proposed?

Add your creative contribution

0 / 200

Added via the text editor

Sign up or


Guest sign up

* Indicates a required field

By using this platform you agree to our terms of service and privacy policy.

General comments