Google has recently announced the development of a generative AI-powered virtual try-on feature for clothes. This innovative tool aims to recreate the in-store shopping experience by allowing users to preview clothing items on real models before purchasing them online.
The revolutionary feature will be integrated into Google Shopping and will utilize a diffusion-based model developed in-house. This model - which is also used in the cutting-edge text-to-art generators Stable Diffusion and DALL-E 2 - learns by gradually subtracting noise from a starting image filled with noise and moving it closer to a target image.
Using a vast number of image pairs depicting individuals wearing garments in various poses, Google trained its model to accurately drape, fold, stretch, and display clothes on a selected lineup of lifelike fashion models. As a result, users are provided with a more realistic visual representation of the clothing items and an enhanced online shopping experience.
This groundbreaking virtual try-on feature is now available to U.S. shoppers using Google Shopping for women's tops from brands such as Anthropologie, Everlane, H&M, and LOFT. A "Try On" badge will be visible in Google Search, while men's tops are scheduled to be introduced later this year.
As Lilian Rincon, senior director of consumer shopping product at Google, pointed out, customer satisfaction should remain paramount during online shopping experiences. In fact, recent surveys have revealed that 42% of online shoppers do not feel represented by images of models, whereas 59% feel dissatisfied with an item due to unforeseen discrepancies when trying them on.
Although virtual try-on technology is not an entirely new concept, with companies like Amazon, Adobe, and Walmart having previously experimented with similar technologies, Google's AI-powered approach is set to revolutionize the online shopping industry.
Alongside the virtual try-on feature, Google is also launching AI-powered filtering options for clothing searches and visual matching algorithms in its Shopping platform. These filters will provide users with the ability to narrow down their searches across various stores, considering factors like color, style, and pattern. This will further streamline the online shopping experience and aid customers in finding the perfect garments tailored to their preferences.
Such advancements in the world of no-code and low-code technologies, like the ones seen on AppMaster's platform, undoubtedly benefit online businesses by offering cost-effective and efficient solutions. By leveraging the power of next-generation tools like generative AI, more advanced applications are being created, simplifying the online experience for users worldwide.
