The new feature has already been rolled out with a limited number of brands currently supporting it.
Google has launched a new online shopping feature meant to help consumers to better understand how clothing will fit them.
The feature is immediately available for all users, though only a limited number of brands support it.
Among the brands that support the new online shopping feature include LOFT, H&M, Everlane, and Anthropologie, among a few others. It has been rolled out in Google’s Shopping tab and makes it possible for users to see how an item of clothing would fit on a body type like theirs.
To work, the feature employs a diffusion-based image generation model. Through that technology, it produces a realistic version of the item of clothing being considered by the consumer. It uses the clothing item to dress one of 40 different models that range from XXS to 4XL in size. This way, a consumer will be better able to determine how a specific piece of clothing and size will fit a body type like their own.
The new online shopping feature currently works only with women’s tops, but Google will expand it.
At launch, the virtual try-on feature from Google will work only with women’s tops, among the brands that support it. That said, Google is working to expand to more clothing types as well as additional brands and categories as time goes on.
When using the Shopping tab to browse women’s tops, consumers now have a “try on models” option at the first image of the item, if that item is eligible for the use of the new feature.
The feature is based on the Imagen model from Google, which uses AI to send the top’s image onto different sized models in two separate neural networks. From there, those neural networks collaborate to build a version of the clothing item that will fit the body shape and pose of the model. From there, the AI goes over the way clothes will typically fold, drape, stretch, or wrinkle, to create a final try-on step that will provide useful insight to the online shopping consumer.
