Google Labs, Google's experimental arm, is Testing a brand new image generator called Whisk. This tool allows users to enter images as an alternative of text, allowing them to remix a photograph by changing the topic, scene, and magnificence.
Whisk uses Google's Imagen 3 image generation model to mix three images: one for the topic, one for the scene, and one for the style. For example, you possibly can select a photograph of yourself as the topic, a futuristic landscape because the scene, and an anime style for the ultimate look.
The model mechanically generates an in depth caption of your images, which then helps Imagen 3 create a remix of the photo. You can even enter text prompts to further define the specified end result, including detailed descriptions reminiscent of “Subject rides a flying bike.”
Because Whisk only focuses on a couple of key features of every image, the corporate explains that the outcomes may not all the time be what you expect. For example, the generated motif could differ in size, weight, hairstyle or skin tone. Google says you possibly can view and edit the underlying prompts at any time.
The experiment is currently only available to users based within the United States labs.google/whisk.
