Google has unveiled its new virtual try-on feature for clothes shopping on Search, making it easier for shoppers to buy clothes online. The feature uses generative AI to create highly detailed portrayals of clothing on real models, with different body shapes and sizes. Unlike many other virtual try-on tools in the market, Google’s new method produces photorealistic images of people wearing the outfit rather than dressed-up avatars.
New Diffusion-Based AI Model
To produce accurate images of clothing on different body shapes and sizes, Google developed a new diffusion-based AI model. The model was trained using Google’s Shopping Graph, a comprehensive dataset of products and sellers, on millions of image pairs showing different people wearing an outfit in two different poses. The model learned to render outfits on images of different people standing in different poses, whether sideways or forward. This way, whenever a user exploring an outfit on Search hits the try-on button, they can select a model with a similar body shape and size and see how the outfit would fit them.
How the Feature Works
The new try-on feature on Google Search works by selecting a garment and model image as input data. Each image is sent to its own neural network and shares information with other networks in a process called “cross-attention” to generate the output: a photorealistic image of the person wearing the garment. The feature works only for women’s tops from brands across Google at the moment, but as the training data grows and the model expands, it will cover more brands and items.
Benefits of the New Feature
The new feature on Google Search addresses the fitting errors of many virtual try-on tools in the market. Instead of deforming a clothing image to fit a person’s avatar, Google’s generative AI model can accurately reflect how the clothing would drape, fold, cling, stretch, and form wrinkles and shadows on a diverse set of real models in various poses. The new feature will make online shopping a tad easier for shoppers in the U.S. and will soon be available for men’s clothing later this year.
Google’s new virtual try-on feature for clothes shopping on Search is a game-changer for online shopping. By using generative AI and producing highly detailed portrayals of clothing on real models, the feature addresses the fitting errors of many virtual try-on tools in the market. As the training data grows and the model expands, it will cover more brands and items, making online shopping more accessible and convenient for millions of shoppers.
Leave a Reply