Google launched Doppl, an experimental AI-powered application for iOS and Android, allowing users to virtually try on outfits using a digital representation of themselves.
The Doppl application operates by requiring users to upload a full-body photograph of themselves. This uploaded image is then utilized by the application to generate a digital version of the user. Subsequent to the creation of this digital representation, users can input photographs or screenshots of various outfits. These outfit images can originate from diverse sources, including but not limited to, personal photographs of clothing observed in retail environments such as thrift stores, images of apparel worn by acquaintances, or screenshots captured from social media platforms. Upon the selection and input of an outfit image, Doppl processes this information to produce an image depicting the user’s digital representation attired in the chosen outfit.
Doppl incorporates the capability to convert the static images of the user’s digital representation wearing selected outfits into AI-generated videos. This functionality is designed to provide users with a more dynamic visualization of how the attire might appear. Users have the option to save these visualized looks, enabling them to maintain a collection of their virtual try-ons. The application also facilitates the sharing of these virtual appearances with other individuals. This sharing feature allows users to transmit their visualized outfits to contacts or across platforms.
The technological foundation of Doppl extends from previously established capabilities within Google Shopping, which permit the virtual try-on of garments. Google asserts that by releasing this functionality as a distinct, standalone application, the company aims to enhance accessibility to the feature. This strategic decision is predicated on the belief that a dedicated application streamlines the user experience and offers an engaging method for individuals to explore personal style preferences.
Google’s Gemini now lives in your terminal
The company also anticipates that this standalone application may facilitate the collection of additional data regarding the operational dynamics of such applications. This data acquisition is intended to inform and assist future developmental efforts by Google within the domain of virtual try-on technology. While Google had offered virtual try-on technology prior to Doppl, earlier iterations of this technology primarily focused on presenting clothing items on a diverse array of models with varying body types. In contrast, Doppl distinguishes itself by enabling users to apply clothing to an animated depiction of their own body.
Google communicated its aspirations for Doppl in a blog post, stating, “We hope Doppl helps you explore your style in new and exciting ways.” The company also provided a clarification regarding the developmental stage and potential limitations of the application. Google categorized Doppl as a “Google Labs experiment,” indicating its early developmental phase. Consequently, the company acknowledged that the application “might not always get things right.” Specific areas where inaccuracies may occur include “Fit, appearance and clothing details,” which “may not always be accurate.”
