US shoppers can now virtually try on apparel from stores across the web and even see how it would drape, fold, cling, stretch, form wrinkles on a diverse set of real models in various poses, different skin tones, body shapes, ethnicities, hair types, price points, colour, style and pattern.
- Google’s new AI model powers its virtual try-on (VTO) feature through generative AI model that uses a technique called diffusion to show what clothes look like on a wide range of people with different body shapes, sizes, and poses.
- This includes those subtle but crucial details like how something drapes, folds, clings, stretches and wrinkles.
- The guided refinements can also help shoppers fine-tune products until the perfect piece is found.
- Unlike shopping in a store, buyers are not limited to one retailer: They will be able to see options from stores across the web. This feature will be available for tops to begin with, right within product listings.
Working with Data: Working alongside its Shopping Graph, the world’s most comprehensive data set of products and sellers, this technology can scale to more brands and items over time.
- The ML-powered Shopping Graph houses more than 35 billion product listings.
How It Works: The new regenerative AI model selects people ranging in sizes XXS-4XL representing different skin tones (using the Monk Skin Tone Scale as a guide), body shapes, ethnicities and hair types.
- If there’s an image of someone wearing a shirt standing sideways and another of them standing forward, the AI model learns to match the shape of the shirt in the sideways pose with the person in the forward pose, and vice versa, until it can generate realistic images of the shirt on that person from all angles.
- To take it up a notch, Google repeated this process using millions of random image pairs of different garments and people. The result allows shoppers to see what a top looks like on the model of their choice.
- Shoppers need to tap products with the “Try On” badge on Search and select the model that resonates most with them.
