New entrant in the field of generative artificial intelligence with Whisk, a new product from Google. The company says its tool is intended for quick visual exploration rather than precise edits: the results aren't perfect, but have editing features to correct inaccuracies.
AIs prefer Whisk-ASS?
Whisk is a recent experiment from Google Labs that lets users create images using other images as starting points. By combining the Imagen 3 model with Gemini's visual understanding capabilities, Whisk generates unique images based on user-provided topics, scenes and styles. This tool makes it easy to visualize and recompose ideas in innovative ways.
If the user does not have images, Whisk can offer them generated by AI – conclusion, the pitfalls of copyright are very moderately avoided. The user will also add additional textual details if desired.
Don't you want a Whisk-Y first?
Once the prompts are provided, Whisk generates corresponding images and text descriptions for each, allowing users to save, download, or refine them by adding more text or editing the image directly.
Obviously, Whiksi relies on Gemini which will produce a description of the images, before propelling them into Imagen 3, the image generation tool. “ In the background, the Gemini model automatically generates a detailed description of the provided images, which is then processed by Google's newest image generation model, Imagen 3. This process captures the essence of the provided images, allowing to remix them in a creative way », Indicates the Mountain View company.
READ – Readers prefer the poetry of Artificial Intelligence
This solution is designed for rapid visual exploration, rather than precise pixel-level edits. It aims to encourage experimentation and creativity by allowing users to quickly test different ideas and download the ones they like.
Whisk uses Google's Imagen 3 image generation model. To test it, it will be at this address, while waiting for your turn, since the tool is only available in the United States for the moment. Note that Google will soon distinguish images generated or modified by AI in its search results, clearly indicating their origin.
READ – Future of publishing: an AI tool that assesses the quality of books
Image credits: Google, Imagen 3
By Clément Solym
Contact : [email protected]
