Google has launched a new AI image generation tool called Whisk, which completely changes the traditional way of image generation. Unlike the previous need for lengthy text descriptions, Whisk allows users to directly upload images as prompt words to specify the theme, scene and style of the generated images. Users can provide multiple pictures for each of these three aspects, and Whisk will cleverly integrate these visual information to generate unexpected images. Even if there is no suitable picture, Whisk also provides a random picture generation function to help users quickly start creation.
Google has released a new AI tool called Whisk, which subverts the traditional image generation model and allows users to use pictures as prompt words instead of lengthy text descriptions.
With Whisk, you can upload images to specify the theme, scene, and style of the AI-generated images, and you can use multiple images for each of the three.
For example, I entered a picture of a pig and a picture of a cat, selected the style of an illustration, and then did not enter any prompt words (of course, you can also choose to use text prompts at the same time.), it automatically gave me Such a rendering was generated. Note that SCENE generally inputs scene pictures. Of course, it is no problem if you have to input character pictures like me. It will be automatically fused and generated. Sometimes it may not be very relevant, but there will be unexpected surprises.
If you don’t have a suitable image on hand, you can click on the dice icon to have Google automatically fill in some images as prompts (these images also appear to be AI-generated).
I clicked casually, and Google randomly gave pictures of a dog, a boat, and embroidery to see what kind of effect the mixture would produce:
The effect is pretty good. The elements of the three pictures are perfectly blended together to form an interesting embroidery picture~
Clicking on the image, I discovered that Whisk also provides text hints for each generated image. If you are satisfied with the result, you can save or download the image; if you want to optimize it further, you can add more text in the text box, or click on the image directly to edit the text prompt.
Google emphasized in a blog post that Whisk is designed for "rapid visual exploration, not pixel-precise editing." The company also says Whisk can "go off track," so it allows users to edit the underlying prompt words.
I played around with the Whisk tool for a few minutes and found it very interesting. While image generation takes a slightly annoying few seconds, and the resulting images can sometimes look a little strange, the iterative process is fun.
Google said Whisk uses the latest version of the Imagen3 image generation model, which was also officially released today. Google also released a new generation video generation model Veo2, which is said to be able to understand the "unique language of movies" and have "fewer" artifacts such as extra fingers. Veo2 will launch first in Google's VideoFX, and users can apply to try it out through a Google Labs waitlist, with plans to expand to YouTube Shorts and other products next year.
All in all, the emergence of Whisk has brought new possibilities to image generation, allowing users to express their creativity through visual means and more conveniently achieve personalized customization of images.
Product experience address: https://top.aibase.com/tool/whisk
With its innovative image generation method, Whisk provides users with a more intuitive and convenient image creation experience, which is worth a try. Hurry up and visit the link to experience this powerful AI tool!