Skip to main content

Google expands generative AI virtual try-on to new product category

Google virtual dress try-on
Google's new AI-based virtual dress try-on.

Google’s generative AI-based virtual try-on experience is moving beyond men’s and women’s tops.

The tech giant, which began integrating new generative AI functionality into its search engine in May 2023 and started virtually showing shoppers tops on a selection of real models in June 2023, is now offering a virtual try-on tool for women’s dresses.

"Shopping online to update your seasonal wardrobe can be fun — and challenging," Lilian RinconVP, product management and Ira Kemelmacher-Shlizerman, senior staff research scientist, shopping, Google said in a corporate blog post. "It’s hard to tell how a new style will look on you, and buying something that doesn’t work can send you through the returns process, which is burdensome for both shoppers and retailers."

Google’s generative AI model can take one clothing image and accurately reflect how it would drape, fold, cling, stretch and form wrinkles and shadows on a diverse set of real models in various poses. The company selected people ranging in sizes XXS-3XL, representing different skin tones, body shapes, ethnicities and hair types.

Since dresses cover more of a person's body than tops, Google developed a new technique called the VTO-UNet Diffusion Transformer (VTO-UDiT) which isolates and preserves a person’s important features to give a more accurate portrayal of the dress as well as the person wearing it.

In the U.S, shoppers can search for dresses on Google Search and click any style that includes a "try on" badge. From there, customers can see what that garment looks like on a diverse set of real models, ranging in sizes from XXS-3XL. Once a customer finds what they’re looking for, they can then click out to the retailer’s site to buy it.

Advertisement - article continues below
Advertisement

In the U.S, shoppers can search for dresses on Google Search and click any style that includes a "try on" badge. From there, customers can see what that garment looks like on a diverse set of real models, ranging in sizes from XXS-3XL. Once a customer finds what they’re looking for, they can then click out to the retailer’s site to buy it.

According to Google data, virtual try-on images on its search engine receive 60% more high-quality views and on average, people try on clothes using four models per product. Google says shoppers are also more likely to visit a brand’s site after viewing virtual try-on images.

Customers can virtually try on dresses from hundreds of dresses across the Google Shopping Graph, a real-time, AI-based dataset of products, inventory and retailers with more than 35 billion listings, in a few taps. Available brands include Simkhai, Boden, Staud, Sandro and Maje.

During September 2024 Fashion Week in New York, hundreds of brands will be showing dresses from their latest collections on runways across the city. Following ready-to-wear brand Simkhai's September Fashion Week show, shoppers will be able to virtually try on select dresses from the collection straight from the runway and pre-order them from the Simkhai site. 

Shoppers can search for Simkhai dresses on Google and look for the “try-on” icon to get started. 

[READ MORE: Google makes images, search results shoppable]

X
This ad will auto-close in 10 seconds