Google is expanding its generative AI initiatives with Whisk, a new experimental tool designed to explore the technology’s creative potential. Following the success of its existing AI tools, the company aims to further enhance content generation capabilities through this latest development.
Whisk introduces a new approach to AI image generation by enabling users to create visuals using image prompts instead of detailed text descriptions. Users can drag and drop images to generate content, making the tool accessible to those without prior experience in AI-generated art.
Generative AI relies on deep-learning models to produce high-quality content based on existing data, but text-based prompts often miss certain design elements. Whisk addresses this limitation by allowing users to customize specific aspects of an image, including the main subject, background, and artistic style.
The tool utilizes Google’s Gemini model to analyze selected images and generate detailed captions. These descriptions are then processed by Imagen 3, Google’s latest image generation model, to create new visuals that reflect the essence of the original subject without producing an exact replica. This method enables users to experiment with different styles and compositions for various creative applications.
Whisk extracts only a limited set of key features from input images, so results may vary from initial expectations. Users can review and modify the underlying AI-generated prompts to refine their outputs and achieve the desired result.
You can download Whisk here: https://labs.google/fx/tools/whisk.
