Google Labs Doppl app enables virtual clothing try-ons

Google Labs has released Doppl, a smartphone app for Android and Apple devices that allows users to try on clothing virtually. It uses artificial intelligence technology to replace the clothing of people in photographs with selected outfits.

Artificial intelligence image processing has been used from automated targeting on the battlefields to the generation of fantastic, never-before-seen art. Once an AI model has been fed millions of images of people wearing various clothing and without, it learns how clothing should fit and drape on a person properly.

Doppl first takes an image of the user as primary input. Although specific algorithms have not been disclosed, the AI uses this image to understand the user's body type, whether they are fat or skinny. Next, images of clothing for virtual try-on are uploaded. The AI also uses these images to understand the type of fabrics used.

Through undisclosed methods, the AI calculates the drape and fit of the clothing to the user’s body, including curves, then generates both photos and motion videos for the user to see how these outfits might fit in reality.

The benefit is that less time and energy are wasted buying and returning clothes that don’t look right or don’t fit well. A potential nefarious use is better identification of people regardless of how they try to cloak themselves from camera surveillance.

The app is available for experimental use from the Google Play Store for Android smartphones and the Apple App Store for Apple smartphones. Readers who wonder what hidden cameras are watching over them when staying in hotels or trying on clothes in fitting rooms might want to use a detector like this one on Amazon.