Google updates his Visual fitting tool powered by artificial intelligenceand the best part is that you no longer need a full-length photo to use it. Thanks to the company's Nano Banana model (part Twins 2.5 Flash Image Suite), the tool can now turn a single selfie into a full-body digital avatar for trying on clothes, and it works in Google Search and Shopping.
All you have to do is upload a selfie and select your usual size; Nano Banana will generate several avatars. Using your face, the AI ​​model generates the rest of your body while maintaining constant proportions. Choose the one you think suits you best and use your avatar to try on shirts, jackets, dresses and more.
Faster and more personalized AI tests
Previously, you had to upload a full-size head-to-toe image, but the new feature saves time that would otherwise be spent searching for a photo or taking it. The new AI-generated avatar does not rely on a predetermined body type or standard shape. Instead, it creates a personalized digital avatar for a more realistic look at fit and style.
The Nano Banana update should result in a richer and more immersive shopping experience, making shoppers more confident in what they're buying before making a payment. If you buy clothes online, upgrading can help you save time, frustration and the hassle of returning an item because you purchased a smaller or larger size.
Google's new shopping feature is currently rolling out in the United States, giving American shoppers the chance to try it out before anyone else in the world. Because this feature is built right into Google Search and Shopping, you don't need to download or sign up for anything extra.
In the future, Google may release additional improvements to its Nano Banana model, creating more realistic avatars and 3D try-ons, as well as deeper integration with apps like Doppl for style and outfit ideas.. I'm a little skeptical about a feature that estimates a person's body type based on their face, but it should still help find a virtual fit and speed up the checkout process.
