What Are Virtual Fitting Rooms and How Do They Work? (2023)
A virtual fitting room is a type of technology that lets shoppers try on items virtually. Shoppers can virtually try on clothing items or...
top of page
Project UMA UC-012
History of the project
The idea of creating a virtual fitting room is not new. For more than 20 years, developers around the world have been trying to do this.
But, despite the constantly moving forward progress in the production of new hardware, processors, application development software, developers in this direction have not advanced a single step, with the exception of some.
We will talk about them in the "Overview" section.
Before starting the project, we studied almost all the developments and trends around the world that existed today, as well as closed projects.
.jpg)
Technologies from the Metaverse
And we figured out what the problem is when creating virtual fitting rooms!
In general, everything ingenious is simple! The main problem of failures is the wrong terms of reference for the developer and the choice of the wrong algorithms for building virtual fitting rooms.
our team
Sergey Dronov - idea, terms of reference and algorithms
Anton Rozhkov - wrote code and algorithms
Anatoly Serov - rendering, 3D graphics
Alexandra Makhova - clothing designer
Elena Nagornaya - designer, technologist
Elena Dronova - investor
VIRTUAL TRYING 3D with AI
Compared to shoes, glasses, watches, hats, virtual 3D clothing fitting is still a challenge. The reason is that the clothes are deformed, taking the shape of the human body. Thus, for a proper AR experience, a deep learning model must determine not only the main key points on the joints of the human body, but also the shape of the body in 3D.
Especially when this body is female, and it is further corrected by underwear.
This is where we started our project, that is, with a virtual fitting room for women's underwear. Perhaps the most difficult thing in digital clothing design.
This is our Uma UC-012
Uma is a super machine with 3D rendering and all the necessary software for creating and processing video infographics.
The visual image of Uma UC-012 came about after overlaying a large number of different scanned bodies of real women.
Uma experiments first on herself before creating Avatars and is constantly learning. She already knows several thousand fabrics, tens of thousands of types of accessories, in general, everything from which products are sewn.
In the future, she plans to sew, pack and deliver goods to the consumer herself.
How we taught Uma
1. Studying the body
*3D scanning of models of different types, composition, age of bodies.
Uma already knows many real bodies of different builds, proportions, skin colors and features, as she has already processed more than one hundred bodies, both without clothes and in them, using 3D photogrammetry.
She knows when clothes are small, large or fit. What is very important, when the future customer starts trying on products for her real sizes, she will show how the body will change.
*Uma learns to change his body according to the given parameters. *Uma compares her pixels on a live model
2. Uma studied together with designer Alexandra Makhova what types of lingerie exist in general, what trends are in fashion today, and even made up her first collection.
.png)
.png)
.png)
.png)
.png)
.png)
3. Together with designer-technologist Elena Nagornaya, Uma studied the technology of building patterns and the technology of direct sewing of linen with subsequent landing on any body.
*This is how virtual tailoring of linen takes place, and not only, in general, any clothes according to patterns, followed by landing first on the body of Uma, and then on a living prototype of the model.
The whole process is visualized and allows the designer, designer and technologist to immediately see the result of the work and allows you to correct the models in real time.
4. And only after such in-depth training did we move on to designing a virtual fitting room UC fitting room
*Virtual fitting room consists of a Universal Communications Terminal (UCt), the role of which is performed by a personal smartphone with pre-installed software, an online store web portal and a developer server (UC.Technology LLC).
Personal cabinet
Server UC.Technology
UMA UC-012
Smartphone
Online store
View 4.1
Personal cabinet
*The client opens a multifunctional personal account on the UC.Technology platform.
In the personal account, the client will see several pre-installed Avatars for different client typologies
On the smartphone, we download the application from the Personal Account on the developer's server (UC.Technology LLC).
After installing the application on your smartphone, you can easily rebuild the Avatar to fit your exact dimensions (View 4.1).
For more demanding clients, we will build an Avatar from four photographs, almost a virtual twin.
Smartphone
*After installing the virtual fitting room application on the smartphone with our Avatar, we go to the website of the online store that has UCt technology and fill the fitting basket on the smartphone (View 4.4).
After completing the fitting, press the AR button and examine the products to the smallest detail right in your interior (View 4.2.), And then if the product looks good on your body and you like it, go to the website of the online store and complete the purchase.
View 4.2
Online store
View 4.4
*On the trading platform of the online store, we install software for online stores.
Next, the store, on its own or with our help, digitizes the products intended for fitting in a virtual fitting room (View 4.3.), And transfers the digitized products to the UC.Technology server, which will immediately appear in the shopping cart (View 4.4).
The client fills the basket for fitting on the smartphone (View 4.4). and after completing the fitting, he goes to the website of the online store and completes the purchase on the terms of the store.
View 4.3
5. Overview of similar technologies
Amazon and Walmart have come closest to our developments and technologies.
Amazon
At Amazon, virtual fitting rooms are handled by a newly created department, formed after buying the start-up Body Labs in October 2017. This company has developed a technology that captures a person from several angles and creates a realistic 3D model of his body, where virtual clothes are then superimposed. In November 2014, a startup from New York received the first $2.2 million to implement its technology, and already in October last year, Amazon, imbued with the idea, bought it in full for $70-$100 million (the exact amount was not disclosed then).
Walmart
Walmart, the world's largest retail chain, has acquired Zeekit, a startup that develops an AI platform that allows shoppers to virtually "try on" clothes while shopping online. The platform operates through a combination of technologies: real-time image processing, computer vision, deep learning and artificial intelligence technologies. "Trying on" clothes for a client, the tool takes into account body parameters, fit, size, and even the material of the clothes themselves.
Zeekit was founded in 2013 by CEO Yael Wiesel, VP Nir Appleboym and CTO Alon Crystal. The main goal of the startup was to reduce the number of returns of clothing purchases made on the Internet thanks to the possibility of virtual fitting. In the past, the company has already had experience of cooperation with some stores and brands, including Macy's, Asos, Tommy Hilfiger, Adidas and others. In 2016, the platform also provided services to designers at New York Fashion Week. The cost of the deal between the trading network and the startup was not disclosed, but according to information from reliable sources, it amounted to $72 million. Zeekit is known to have raised $16 million in investment since 2014.
Learn more about Zeekit and Body Labs tech, as well as deals with Amazon and Walmart, in our blogs below.
UC Technology
The Uma UC-012 project is similar in terms of the declared functionality, but has not yet been implemented, or has been partially implemented, and cardinally differs in the following:
1. The most simple and understandable for the client and the seller of clothing.
2. A whole range of required tasks from a virtual fitting room has been implemented.
3. Does not require additional special knowledge and specialist content from the seller and the client to use the UC.Technology technology
4. Does not require additional computing resources, so everything is implemented inside the server capacities
AI Uma UC-012. On the website of the online store there is only a button "try on with UCmeta technology" and video instructions on how to take measurements (View 4.5)
5. The technology not only allows continuous learning and improvement, but can also be used in the production of garments in fully automatic production without the participation of manual labor.
6. And finally, our technology is already working today!
View 4.5
bottom of page