Google unveils a new feature “Virtual Try-on” for people to look through clothes for different sizes, body types and range of skin tones to ease the online-shopping experience.
E-commerce websites like Amazon, Flipkart has unfolded tons of opportunities and business sales, making it as one of the biggest platforms in the world thereby providing practicality of purchasing and shopping from home. People started to buy everything online, from clothes, phones to groceries and medical items.
Though most of the online-shopped products are convincing for people, people often feel unsatisfied in few categories on their first purchase, leading to returns. Apparels (clothing) section is the foremost in the list, as 59% of online-purchasers feel dissatisfied with an item they shopped for online because it looked different on them than expected.
Addressing the issue, Google brings AI & ML technology by introducing a feature called “Virtual Try-on”.
Virtual Try-On Feature
Ever purchased a shirt that looked attractive to the ‘fit’ model in shopping websites, but looks different on you? Google takes a moment here, with this “Virtual Try-on” (VTO), which shows you how clothes look on a variety of real models at different size, body shape and skin tones or even ethnicities.
New generative-AI model of Google will dress-up people at different sizes XXS – 4XL with range of skin tones, body shapes, hair color/types with the shirt / top / pant / skirt you choose. While viewing clothes in Google, a button showing “Try-on models” will let you choose and see what the apparel would be like for different models, so you can select the one that resonates most with you.
Technology behind Virtual Try-On
Shopping Graph
Google’s artificial intelligence and machine-learning tools assists to make happen the VTO feature. Google had recently developed a machine-learning model called “shopping graph”, which comprises of all the data of apparels and all other products. It houses more than 35 billion product listings. Shopping graph of Google filters out products based on user choices, after comprehending all relevant data from the web – like images, descriptions, reviews, videos and more.
Shopping graph indeed sees through pictures, understanding the patterns or colors of clothes or other products. The machine-learning model could also suggest you what suits the best combo for a chosen apparel, like what trouser would be better for a shirt, or what shoe goes with the outfit.
Further, shopping graph could even elaborate a buying guide for any product – things to consider before purchasing a product. Say for example, if you are looking for an air purifier, a buying guide for that would detail you the type, price range, popular brands, and even the functioning of air-purifier, ultimately to prepare you for both before-shopping and after-shopping experience.
VTO – Creating images with AI
With all data available via Shopping graph, Google then selects different people on sizes varying from XXS to 4XL, range of skin tones, body types, ethnicities and hair types and put every costume individually on each of them using artificial intelligence technology.
Google first gathers pairs of images of different persons wearing a garment in two different poses – say, front pose and side pose. The AI model learns to match the shape of the shirt with the person by Google’s diffusion-based AI model that layers a photo and replaces the pixels of cloths with the new apparel, and generates realistic images of that person wearing the cloth you choose.
It works on many brands including Anthropologie, LOFT, H&M and Everlane and the tool is expected to expand to more brands in coming months. Currently the VTO is restricted only to U.S. people and will be rolling out worldwide in coming months, Google said.
The tool might one day put yourself as a model with your virtual avatar in online trying out clothes, making your phone as a personal virtual trial-room.
Hope you find the page useful!
(For more such interesting informational, technology and innovation stuffs, keep reading The Inner Detail).
Kindly add ‘The Inner Detail’ to your Google News Feed by following us!
