Google’s new AI tool changes images based on user’s text input

Google Imagen Editor AI
Honor X9b Ad
Honor X9b Ad

A company as influential as Google will certainly not sit back when it comes to developing AI generative tools. Hence, just like AI text-to-image generators, such as DALL-E 2, Dream by Wombo, and Stable Diffusion, Google has recently launched a BETA version of its own AI model; Imagen Editor. It is an interesting tool that uses generative AI to recreate pictures in one sentence. If you are looking forward to knowing about it more, this article is specifically written for you!

Google Imagen Editor Overview:

The clever wordplay between “Imagine” and “Image” readily gives off a clear hint about what the Imagen Editor is all about. It is an AI tool that creates customized images from a single sentence prompted by the user. The tool doesn’t generate the entire image content from scratch. Instead, it utilizes millions of images from the internet and modifies these existing images according to the user’s preferences.

Though there are many AI generative tools- Imagen Editor is unlike any other. The dataset on which the Imagen Editor’s AI model is trained is LAION-400M. These sorts of information are rather not disclosed by AI companies, like DALL-E 2, but this time around, it is publically informed.

Capabilities it holds for the time being

This generative tool is at present mainly capable of generating two distinct outputs called City Dreamer and Wobble. City Dreamer allows the creation of unique buildings, while Wobble generates animated creature-like characters based on descriptions. With City Dreamer, users can provide unique descriptions of buildings and receive output with weird or visually appealing designs. Meanwhile, Wobble allows users to enhance the image of animated creatures by adding effects, clothing, or background elements.

Due to its current emphasis on generating specific types of images, Google’s Imagen Editor falls slightly behind in comparison to well-established generative AI tools like DALL-E 2.

You might want to read these:

About Availability

Google typically launches its new AI-related technology through an app named AI Test Kitchen. This time as well, the new Imagen Editor is available on the same platform with beta access. This approach allows Google to gather enough user feedback and address any issues before a public release.

Also, Google itself says, the application is not ready for security-related concerns, making it a wiser decision to introduce a beta version rather than facing huge public backlash on the initial release.

By simply providing instructions in sentences, Imagen enables the addition of realistic-looking elements in any image.

Here’s how to use the beta version

Currently, Google Imagen Editor is only available for beta testing to a very small group of people. To use Imagen, all you need to do is visit the AI Test Kitchen website and register. If you are fortunate enough to gain access to the beta testing, which only a few get, make sure you provide enough feedback to contribute to the development of the system.

Later when Google decides to make it publicly available, it will probably be accessible from Google Photos. You supposedly have to add an image to Google Photos and provide a description of the required changes. Imagen Editor will then generate the modified image according to the instructions. The main purpose of the tool is to make changes to the picture without having to manually edit it.